Effect of Cubic Crystal Morphology on Thermal Characteristics and Mechanical Sensitivity of PYX

Abstract

To investigate the influence of the cubic crystal morphology on the thermal properties and sensitivity of 2,6-bis(picrylamino)-3,5-dinitropyridine (PYX), cubic PYX (CPYX) crystals were prepared using the antisolvent method. Scanning electron microscopy (SEM), laser particle size analysis, X-ray diffraction (XRD) and Fourier transform infrared spectroscopy (FT-IR) were used to characterize the morphology, particle size and structure of the prepared products. The thermal behavior, thermal decomposition kinetics, thermal safety parameters and thermal decomposition mechanism of CPYX were investigated by differential scanning calorimetry–thermogravimetry–mass spectrometry–Fourier transform infrared spectrometry (DSC-TG-MS-FT-IR) and in situ FT-IR experiments. Meanwhile, the mechanical sensitivity of CPYX was determined by means of the explosion probability method. The results showed that the product had a smooth cubic morphology and small crystal aspect ratio with an average particle size (d₅₀) of 10.65 μm, but it had no distinct differences from the crystal structure of raw PYX (RPYX). The thermal decomposition peak temperature, the self-accelerating decomposition temperature and the critical temperature of the thermal explosion of CPYX increased by 7.2 °C, 6.1 °C and 10.4 °C, respectively, compared to RPYX. Similarly, the apparent activation energy increased by 15%. Besides these, the impact sensitivity and friction sensitivity of CPYX decreased by 36% and 20%, respectively, compared to RPYX. The decomposition process of CPYX contains two stages. The first stage involves the breakage of N-H bonds and -NO₂ groups with the release of CO₂, N₂O, NO, HCN and H₂O, followed by the thermal decomposition of the resulting intermediate and the release of CO₂, N₂O and HCN in the second stage.

1. Introduction

Heat-resistant energetic materials (EMs) with excellent thermal stability can endure high-temperature environments for a long time [1,2,3]. In particular, 2,6-Bis(picrylamino)-3,5-dinitropyridine (PYX) is a type of heat-resistant EM, first synthesized in the 1970s [4], that has been widely used in oil-well drilling [5,6]. As shown in Figure 1, PYX has a symmetrical molecular structure with strong intermolecular hydrogen bonds and π-bonds, and its unique structure leads to excellent molecular stability [7,8]. The decomposition temperature of PYX is 360 °C, which is higher than that of hexanylstilbene (HNS) [9]. A vacuum stability test (VST) showed that PYX released 0.49 mL of gaseous products under heating for 140 min at 260 °C, which was lower than that of 1,3,5-triamino-2,4,6-trinitrobenzene (TATB) and 2,6-dimaino-3,5-dinitropyrazine-1-oxid (LLM-105) [10]. Because of its outstanding thermal stability, PYX is expected to replace HNS and become one of the most commonly used heat-resistant EMs in the future [11]. However, the morphology of PYX crystals manufactured by this process is needle-like or rod-like, with a relatively large aspect ratio and rough surface [12], which will seriously affect the process’s performance and the application performance of PYX. It is known that EMs with a small aspect ratio and regular morphology are preferred; these qualities can reduce the sensitivity, improve the dispersion and increase the bulk density, thereby improving the properties of the EMs [13]. Therefore, in order to improve the performance of PYX products, it is crucial to optimize the morphology of PYX.

Crystallization is one of the most important methods to reshape the crystal morphology [14]. Currently, there are several studies on the preparation of submicron PYX crystals [15,16], but the reshaping of the crystal morphology of PYX by crystallization has rarely been documented. Solution crystallization is a common method used to regulate and control the crystal morphology of EMs [17,18]. Unfortunately, PYX is only easily soluble in strongly polar solvents such as dimethyl sulfoxide (DMSO), N,N-dimethyl formamide (DMF) and N-methyl-pyrrolidone (NMP), which can be miscible with many other organic solvents; hence, modifying the morphology of PYX via the antisolvent method is an appropriate choice. Recently, the antisolvent method has been widely used in the crystallization of EMs, such as nitroguanidine (NQ) [19] and LLM-105 [20], with its significant advantages including a low operation temperature, high yield and low setup cost.

For EMs, it is also significant to analyze their safety properties, such as their thermal stability and mechanical sensitivity, which can provide supportive data for their safe storage and application. In this work, the antisolvent method was used to prepare cubic PYX (CPYX) crystals with a small aspect ratio and regular morphology. The morphology, particle size and structure were characterized firstly. The thermal characteristics, including the thermal behavior, thermal decomposition kinetics, thermal safety parameters and thermal decomposition mechanism, of CPYX were investigated in detail. Moreover, the impact sensitivity and friction sensitivity of CPYX were determined by means of the explosion probability method. Through the properties of CPYX compared to raw PYX (RPYX), the effect of the cubic morphology on the thermal stability and mechanical sensitivity of PYX crystals were analyzed.

2. Materials and Methods

2.1. Chemicals

RPYX was provided by the Xi’an Modern Chemistry Institute. N-methyl-pyrrolidone (AR grade, abbreviated as NMP) was purchased from the Sinopharm Chemical Reagent Company. Deionized water was obtained by a water purification apparatus (Ulupure, Chengdu, China).

2.2. Antisolvent Crystallization Experiment

NMP and water were selected as the solvent and antisolvent, respectively, and RPYX was crystallized by the antisolvent method. Then, the principle of the preferential dissolution of the particle edges was used to obtain CPYX crystals. The specific experimental steps were as follows. (1) A certain amount of RPYX and NMP was added to a glass crystallizer, and the magnetic stirrer was turned on to dissolve the PYX sufficiently, with a stirring rate of 300 rpm at 60 °C. (2) Deionized water was added to the crystallizer with an addition rate of 2 mL·min⁻¹ to form a supersaturated solution, and then the crystal nucleus were generated and grew. The volume ratio of water and NMP was 5:1. (3) When the crystallization process was over, the temperature was adjusted to 65 °C and held for 30 min, to slowly dissolve the particle edges. (4) The resulting precipitates were filtered and washed several times with ethyl alcohol and then dried in a drying oven at 80 °C for 90 min to obtain CPYX crystals.

2.3. Characterization

The powder X-ray diffraction pattern was collected by a X-ray diffractometer (XRD, DMAX2400, Rigaku, Tokyo, Japan). The scanning range was 5°~50° with Cu-Kα radiation generated at 40 kV and 30 mA. The Fourier transform infrared spectrum was collected by a Fourier transform infrared spectrometer (FT-IR, TENSOR27, Bruker, Karlsruhe, Germany). The spectral range was 4000~400 cm⁻¹.

The morphology was characterized by a scanning electron microscope (SEM, Quanta600, FEI, Hillsboro, OR, USA). The acceleration voltage was 10 kV and the emission current was 5 μA. The particle size distribution was characterized by a laser particle size analyzer (Master Sizer, Malvern, UK). The particle size range was 0.02~2000 μm and deionized water was used as a dispersant.

The thermal decomposition property was characterized by DSC-TG-MS-FT-IR and in situ FT-IR experiments. For the differential scanning calorimeter and thermogravimeter (DSC-TG, STA449F3, Netzsch, Selb, Germany), the temperature range was 30 °C~500 °C; the heating rates were 5 °C·min⁻¹, 10 °C·min⁻¹, 15 °C·min⁻¹ and 20 °C·min⁻¹; the sample mass was 0.4~0.6 mg; and the measurements were performed under a dynamic atmosphere of nitrogen at a flow rate of 50 mL·min⁻¹. For the mass spectrometer (MS, QMS403, Netzsch, Selb, Germany), the resolution was <0.5 amu and the detection limit was >1 µg·g⁻¹. For the Fourier transform infrared spectrometer (FT-IR, NicoletiS20, Nicolet, Madison, WI, USA), the spectral range was 4000~500 cm⁻¹. For the in situ Fourier transform infrared spectrometer (in situ FT-IR, NEXUS870, Thermo-Fisher, Waltham, MA, USA), the sample was pretreated with KBr, the sample mass was 0.6 mg, the heating rate of the variable-temperature reaction cell was 10·°C·min⁻¹ with a detection temperature range of 25~465 °C, the data collection rate was 60 scans·min⁻¹, the IR resolution was 4 cm⁻¹ and the scanning frequency of the atlas was 8 times per sheet.

The mechanical sensitivity was determined by means of the explosion probability method. For the impact sensitivity test, the drop hammer was 10 kg, the drop height was 25 cm and the sample mass was (50 ± 1) mg. For the friction sensitivity test, the swing angle was 90° and the sample mass was (20 ± 1) mg.

3. Results and Discussion

3.1. Morphology and Particle Size

Figure 2 shows the SEM images of the PYX crystals before and after crystallization. The morphologies of the RPYX crystals and obtained PYX crystals show obvious dissimilarity, suggesting that the antisolvent crystallization process has a significant effect on the morphology of the PYX crystals. Before crystallization, the RPYX crystals are needle-like or rod-like with a large crystal aspect ratio (Figure 2a,b). After crystallization, it is observed that the morphology of the PYX crystals tends to be cubic, with a relatively small crystal aspect ratio (Figure 2c–f). The change in morphology can be mainly attributed to the interactions between the solvent molecules and crystal planes. The crystal growth from the needle direction is inhibited and cubic crystals are formed. However, when the crystallization process is completed without using the principle of the preferential dissolution of the particle edges to raise the temperature, the CPYX crystals still have conspicuous edges and corners (Figure 2c,d), which can cause them to easily form hot spots in the process and application. When raising the temperature slightly after the crystallization process, it is found that the edges and corners of the CPYX crystals (Figure 2e,f) become smooth. This can be explained by the fact that, during the dissolution process, with the increase in temperature, the sample will start to dissolve at the location where the radius of curvature is small, resulting in the preferential dissolution of the particle edges and corners. As shown in

Table 1. Particle size results for RPYX and CPYX.

Sample	d10/µm	d50/µm	d90/µm	N	S/(m2·g−1)
RPYX	26.21	66.90	152.16	0.94	0.17
CPYX	3.54	10.65	21.66	0.85	1.39

d₁₀, d₅₀ and d₉₀ represent the particle size corresponding to 10%, 50% and 90% of the cumulative volume of the particle size distribution; N is the particle size span; and S is the specific surface area.

and Figure 3, the average particle size (d₅₀) of CPYX (Figure 2e,f) is 10.65 μm, which is significantly reduced compared to RPYX. The small particle size of CPYX also greatly increases the specific surface area to 1.39 m²·g⁻¹. In addition, the value of the particle size span of CPYX is 0.85, which is smaller than that of RPYX, indicating that CPYX has better uniformity in its particle size distribution.

3.2. Structure

The crystal structure of PYX was analyzed using the XRD patterns. As shown in Figure 4, the diffraction peaks of CPYX are located mainly at 2θ = 12.3°, 13.9°, 16.2°, 18.4°, 20.8°, 23.0°, 25.4°, 27.6°, 29.6° and 32.5°. It is found that the diffraction peaks of CPYX are consistent with those of RPYX, indicating that the crystal structure of CPYX does not change under crystallization. FT-IR spectroscopy was adopted to further characterize the PYX structure. As shown in Figure 5, the FT-IR spectra of RPYX and CPYX show the same absorption bands of -NO₂ and -NH₂ functional groups in the wavenumber ranges of 1423–1635 cm⁻¹ and 3259–3442 cm⁻¹, respectively, with numerous peaks in the fingerprint region [20,21], suggesting that no polymorphic changes occurred during the crystallization process.

3.3. Thermal Characteristics

3.3.1. Thermal Decomposition Behavior

The DSC-TG curves of CPYX and RPYX at the heating rate of 10 °C·min⁻¹ are shown in Figure 6. As can be seen, the DSC curves show that the thermal decomposition of PYX is divided into two exothermic decomposition stages but there is no melting point. This is identical to the TG results, indicating that the thermal decomposition process is carried out in a solid phase. As for RPYX, the TG curve shows that the first decomposition stage begins at around 341.3 °C and is completed at 384.9 °C, accompanied by 43.0% mass loss. The second decomposition stage is completed at 462.8 °C, accompanied by 22.5% mass loss. The DSC curve shows that the peak temperatures of the first and second decomposition stages are 368.8 °C and 411.8 °C, respectively. As for CPYX, the TG curve shows that the first decomposition stage begins at around 350.8 °C and is completed at 386.0 °C, accompanied by 40.8% mass loss. The second decomposition stage is completed at 463.3 °C, accompanied by 23.6% mass loss. The DSC curve shows that the peak temperatures of the first and second decomposition stages are 376.0 °C and 416.6 °C, respectively. In comparison, the mass loss of RPYX and CPYX is similar, but the peak temperatures of the two decomposition stages of CPYX increase by 7.2 °C and 4.8 °C compared to those of RPYX, respectively. The results indicate that CPYX is more difficult to decompose under thermal stimulation.

3.3.2. Thermal Decomposition Kinetics

The thermal decomposition kinetics of RPYX and CPYX were investigated by the non-isothermal dynamic method. Usually, at least four peak temperatures of thermal decomposition at different heating rates are required, which are commonly determined by DSC or derivative thermogravimetry (DTG). In this study, we used DSC to obtain the peak temperatures of RPYX and CPYX decomposed at four heating rates. As shown in Figure 7, at the heating rates of 5 °C·min⁻¹, 10 °C·min⁻¹, 15 °C·min⁻¹ and 20 °C·min⁻¹, the peak temperatures of CPYX are 368.9 °C, 376.0 °C, 379.7 °C and 381.4 °C, respectively, which are all higher than those of RPYX. Moreover, it is found that the peak temperature increases with the increase in the heating rate. This is because increasing the heating rate causes the decomposition reaction to lag; the greater the heating rate, the greater the reaction’s lag, resulting in a shift in the macroscopic peak temperature to a high temperature [21,22].

Based on the DSC data, the Kissinger [23], Ozawa [24] and Friedman [25,26] methods were used to calculate the kinetic parameters, as shown in Equations (1)~(3). The basic kinetic data of the Kissinger and Ozawa methods are listed in

Table 2. Basic kinetic data of RPYX and CPYX.

Sample	β/(K·min−1)	Te/K	Tp/K	103·Tp−1/K−1	Kissinger Method	Ozawa Method
					ln(β/Tp2)/(min−1·K−1)	lgβ/(K·min−1)
RPYX	5	623.0	633.1	1.5795	−11.2918	0.6990
	10	630.1	641.8	1.5581	−10.6260	1.0000
	15	635.7	646.3	1.5473	−10.2345	1.1761
	20	640.0	651.4	1.5352	−9.9625	1.3010
CPYX	5	630.7	641.9	1.5579	−11.3194	0.6990
	10	635.5	649.0	1.5408	−10.6483	1.0000
	15	638.8	652.7	1.5321	−10.2542	1.1761
	20	644.7	654.4	1.5281	−9.9717	1.3010

T_e is the extrapolated onset temperature of the first decomposition stage; T_p is the peak temperature of the first decomposition stage.

. The kinetic parameters obtained by the three methods are listed in

Table 3. Kinetic parameters of RPYX and CPYX.

Sample	Kissinger Method			Ozawa Method		Friedman Method
	EaK/(kJ·mol−1)	lgA/s−1	rk2	EaO/(kJ·mol−1)	ro2	EaF/(kJ·mol−1)
RPYX	257.16	19.15	0.9908	251.66	0.9913	273.91
CPYX	365.91	27.93	0.9944	358.20	0.9953	315.14

E_aK is the apparent activation energy obtained by the Kissinger method, E_aO is the apparent activation energy obtained by the Ozawa method, E_aF is the average apparent activation energy obtained by the Friedman method (0.2 > α > 0.8), A is the pre-exponential constant and r_k and r_o are the linear correction coefficients.

. The reaction rate curves are shown in Figure 8. The apparent activation energy at different conversation rates, obtained by the Friedman method, is shown in Figure 9. The values of the apparent activation energy of CPYX, calculated by the Kissinger, Ozawa and Friedman methods, are 365.91 kJ·mol⁻¹, 358.20 kJ·mol⁻¹ and 315.14 kJ·mol⁻¹, respectively. From Figure 9, we can see that at a wide range of conversation rates (0.2 > α > 0.8), the apparent activation energy of CPYX shows an increasing trend from 299.68 kJ·mol⁻¹ to 331.95 kJ·mol⁻¹. Moreover, the error increases with the increase in the conversation rate. Compared with RPYX, the apparent activation energy of CPYX, calculated by the three methods, is increased, in which the average apparent activation energy obtained by the Friedman method increases by 15%, indicating that CPYX has a higher thermal decomposition energy barrier.

$$\ln (\frac{\beta }{T_{\mathrm{p}}^2})=\ln (\frac{AR}{E_{\mathrm{a}}})–\frac{E_{\mathrm{a}}}{R{T}_{\mathrm{P}}}$$

(1)

$$\lg \beta =\lg \left(\frac{A{E}_{\mathrm{a}}}{RG\left(\alpha \right)}\right)–2.315–0.4567\frac{E_{\mathrm{a}}}{R{T}_{\mathrm{P}}}$$

(2)

$$\ln (\frac{{\mathrm{d}}_{\alpha }}{{\mathrm{d}}_{\mathrm{t}}})=\ln (Af(\alpha ))–\frac{E_{\mathrm{a}}}{RT}$$

(3)

where β is the heating rate, K·min⁻¹; T_p is the peak temperature of decomposition, K; A is the pre-exponential constant, s⁻¹; R is the gas constant, 8.314 J·mol⁻¹·K⁻¹; E_a is the apparent activation energy, J·mol⁻¹; G(α) is the integral form of the mechanism function; α is the conversion rate; and f(α) is the differential form of the mechanism function.

The thermodynamic parameters, including the activation entropy (ΔS^≠), activation enthalpy (ΔH^≠) and activation Gibbs free energy (ΔG^≠), of the thermal decomposition process can be calculated by Equations (4)~(7). As listed in

Table 4. Thermodynamic parameters of RPYX and CPYX.

Sample	Te0/K	Tp0/K	ΔS≠/(J·mol−1·K−1)	ΔH≠/(kJ·mol−1)	ΔG≠/(kJ·mol−1)
RPYX	614.2	615.4	107.35	252.04	185.98
CPYX	620.3	630.0	261.46	360.67	195.95

, the values of the activation entropy (ΔS^≠), activation enthalpy (ΔH^≠) and activation Gibbs free energy (ΔG^≠) for CPYX are 261.46 J·mol⁻¹·K⁻¹, 360.67 kJ·mol⁻¹ and 195.95 kJ·mol⁻¹, respectively. The larger value of the activation entropy (ΔS^≠) of CPYX indicates a higher degree of disorder compared to RPYX. The positive value of the activation Gibbs free energy (ΔG^≠) proves that the thermal decomposition process of PYX is a nonspontaneous process.

$$A \exp (\frac{-E_{\mathrm{a}}}{R{T}_{\mathrm{p}0}})=\frac{{k_{\mathrm{B}}T}_{\mathrm{p}0}}{h}\exp (\frac{-{\Delta G}^{\ne }}{R{T}_{\mathrm{p}0}})$$

(4)

$${\Delta H}^{\ne }=E_{\mathrm{a}}–R{T}_{\mathrm{p}0}$$

(5)

$${\Delta G}^{\ne }={\Delta H}^{\ne }–T_{\mathrm{p}0}{\Delta S}^{\ne }$$

(6)

$${\mathrm{T}}_{(\mathrm{pi} \mathrm{or} \mathrm{ei})}={\mathrm{T}}_{(\mathrm{p}0 \mathrm{or} \mathrm{e}0)}+b{\beta }_{\mathrm{i}}+c{\beta }_{\mathrm{i}}^2+d{\beta }_{\mathrm{i}}^3$$

(7)

where k_B is the Boltzmann constant, 1.3807 × 10⁻²³ J·K⁻¹; h is the Planck constant, 6.626 × 10⁻³⁴ J·s; T_pi is the decomposition peak temperature, K; T_p0 is the peak temperature when the heating rate tends to be 0, K; T_ei is the extrapolated onset temperature, K; and T_e0 is the extrapolated onset temperature when the heating rate tends to be 0, K.

3.3.3. Thermal Safety Parameters

The self-accelerating decomposition temperature (T_SADT) is defined as the lowest temperature that causes self-accelerating decomposition. The critical temperature of thermal explosion (T_b) is defined as the lowest ambient temperature that causes a thermal explosion. T_SADT, T_b and T_p are important parameters for the safe storage and application of EMs. T_SADT and T_b can be calculated by Equations (8) and (9) [27,28]. As listed in

Table 5. T_SADT and T_b of RPYX and CPYX.

Sample	Tp/°C	TSADT/°C	Tb/°C
RPYX	368.8	341.2	350.2
CPYX	376.0	347.3	360.6

, the T_SADT, T_b and T_p of CPYX are 347.3 °C, 360.6 °C and 376.0 °C, which increase by 6.1 °C, 10.4 °C and 7.2 °C compared to RPYX, respectively, indicating that CPYX has better thermal stability. In general, EMs with a small particle size are easier to decompose compared to EMs with a large particle size because they are more conducive to heat transfer. However, CPYX, with a small particle size, still has better thermal stability compared to RPYX, proving that the cubic morphology has a significant influence on the thermal stability of PYX crystals. This can be explained by the fact that the morphology of CPYX crystals tends to be regular, with smooth crystal edges and few defects, which can reduce the decomposition hot spots under thermal stimulation, thereby resulting in better thermal stability.

$$T_{\mathrm{SADT}}=T_{\mathrm{e}0}$$

(8)

$$T_{\mathrm{b}}=\frac{E_{\mathrm{a}}–\sqrt{E_{\mathrm{a}}^2–4R{E}_{\mathrm{a}}{T}_{\mathrm{p}0}}}{2R}$$

(9)

3.3.4. Thermal Decomposition Mechanism of CPYX

The FT-IR spectra of the gaseous products of CPYX are shown in Figure 10. As can be seen, the small absorbance at the temperature of 371 °C indicates that the decomposition occurs at an early stage. The absorbance achieves a larger value at the temperatures of 384 °C and 427 °C, corresponding to the first and second decomposition stages, respectively. For the first decomposition stage, the main gaseous products include CO₂ (2356 and 671 cm⁻¹), N₂O (2201 cm⁻¹and 2240 cm⁻¹), NO (1843 cm⁻¹and 1923 cm⁻¹), HCN (720 cm⁻¹) and H₂O (3511 cm⁻¹~3824 cm⁻¹) [29]. For the second decomposition stage, the absorption peaks of NO and H₂O disappear, indicating that there may be no NO and H₂O produced. After 453 °C, the gaseous products decrease rapidly. The MS spectra of the gaseous products of CPYX in Figure 11 show that CO₂/N₂O, NO, HCN and H₂O correspond to m/z = 44, 30, 27 and 18. Taking the MS and FT-IR spectra together, the gaseous products produced by the thermal decomposition of CPYX mainly include CO₂, N₂O, NO, HCN and H₂O. Among them, CO₂, N₂O and HCN exist in the two decomposition stages, while NO and H₂O only exist in the first decomposition stage, which is consistent with the FT-IR results. In addition, it can be observed that before the first decomposition stage, a small amount of NO appears first.

The FT-IR spectra of the condensed phase materials of CPYX are shown in Figure 12. The absorption peaks of CPYX before decomposition indicate that it has characteristic functional groups including N–H bonds (3270 cm⁻¹), Ar rings (pyridine ring and benzene ring, 1440 cm⁻¹~1637 cm⁻¹) and –NO₂ groups (1341 cm⁻¹ and 1540 cm⁻¹). When the temperature reaches 362 °C, the absorbance of N–H bonds, Ar rings and -NO₂ groups in the condensed phase decreases with the increase in temperature. At 374 °C, the absorption peaks of the N–H bonds disappear, indicating that the N-H bonds have been broken. Meanwhile, the absorbance of the –NO₂ groups continues to decrease. At 389 °C, the absorption peaks of the –NO₂ groups almost disappear, suggesting that the –NO₂ groups on the Ar rings have been broken, and the first decomposition stage has finished according to the previous DSC-TG results. When the temperature reaches 414 °C, the absorbance of functional groups is further decreased, corresponding to the second decomposition stage.

Based on the above results, the thermal decomposition mechanism of CPYX can be extrapolated, as shown in Figure 13. For the first decomposition stage, the first appearance of a small amount of NO indicates that the –NO₂ groups break down in the condensed phase. Because the –NO₂ groups are present on benzene rings, which, in the para position with N–H bonds, are subjected to the relatively small conjugate action of the molecular structure, they may break down preferentially. Furthermore, the N–H bonds in the condensed phase disappear first, showing the preferential breakdown of the N-H bonds. Therefore, the fracture of the N–H bonds and the breakdown of the –NO₂ groups on the benzene rings in the para position with N–H bonds result in the release of NO and the formation of an intermediate (a) [30]. Then, the decomposition of the benzene rings and the breakdown of other –NO₂ groups on the Ar rings cause the intermediate (b) to form, with five membered rings. Meanwhile, CO₂, N₂O, NO, HCN and H₂O are released. In the second decomposition stage, the intermediate (b) decomposes to generate CO₂, N₂O and HCN.

3.4. Mechanical Sensitivity

The mechanical sensitivity of RPYX and CPYX was determined by the explosion probability method. As listed in

Table 6. Mechanical sensitivity of RPYX and CPYX.

Sample	Impact Sensitivity	Friction Sensitivity
RPYX	76%	64%
CPYX	40%	44%

, the impact sensitivity and friction sensitivity of CPYX are 40% and 44%, which decrease by 36% and 20% compared to RPYX, respectively, indicating that the CPYX has lower mechanical sensitivity. This is because RPYX crystals, with a needle-like or rod-like morphology, are easily broken under impact action or friction action, and many hot spots are formed, leading to high impact sensitivity. After crystallization, the crystal aspect ratio of CPYX is greatly reduced, and the crystals do not easily break or form hot spots under impact action, so that CPYX has low impact sensitivity. In addition, CPYX has smooth crystal edges and few crystal deficiencies compared to RPYX, which means that it does not easily generate hot spots under friction action, thereby resulting in low friction sensitivity.

4. Conclusions

Regular CPYX crystals with an average particle size (d₅₀) of 10.65 μm were prepared by the antisolvent method. The crystal structure of CPYX does not change during the crystallization process. Compared with RPYX, the thermal decomposition peak temperature, the self-accelerating decomposition temperature and the critical temperature of the thermal explosion of CPYX increase by 7.2 °C, 6.1 °C and 10.4 °C, respectively; the apparent activation energy of CPYX increases by 15%; and the impact sensitivity and friction sensitivity of CPYX decrease by 36% and 20%, respectively. The thermal decomposition of CPYX can be divided into two stages. The first stage involves the breakdown of N–H bonds and –NO₂ groups with the release of CO₂, N₂O, NO, HCN and H₂O, and the second stage mainly involves the thermal decomposition of the resulting intermediate and the release of CO₂, N₂O and HCN. The results indicate that PYX, with a cubic morphology, has better thermal stability and lower mechanical sensitivity.