Investigating the kinetics of the thermolysis of 3-nitro-2,4-dihydro-3H-1,2,4-triazol-5-one (NTO) and reduced size NTO in the presence of cobalt ferrite additive

The co-precipitation technique was used to successfully produce nanometer cobalt ferrite (CoF). Utilizing the solvent-antisolvent approach, the reduction in the size of sensitive HEM 3-nitro-2,4-dihydro-3H-1,2,4-triazol-5-one (NTO) was effectively achieved. Using simultaneous thermal analysis, the effect of 5% by mass CoF on the thermolysis of NTO and nanosize NTO (r-NTO) was investigated. The kinetic parameter of NTO and r-NTO in the presence of CoF additive was assessed using three isoconversional methods: Flynn–Ozawa–Wall, Kissinger–Akahira–Sunose and Starink. It was found that lowering NTO’s size and adding CoF may both lower the material’s thermal breakdown temperature, with the former dropping it more significantly than the latter. The activation energy of both NTO and r-NTO was raised in the presence of CoF additive.


Introduction
High energetic materials (HEMs) are extensively utilized in propellants, explosives, and pyrotechnics due to extensive heat and gas released during their thermal decomposition. A few of the most widely known HEMs include ammonium perchlorate (AP), cyclotetramethylene tetranitramine (HMX), and cyclotrimethylene trinitramine (RDX). These HEMs are highly sensitive to external stimuli such as shock, and friction; therefore, it is a very challenging task to handle such sensitive HEMs during their handling as a small mistake can lead to an accident [1,2]. More research is required to replacing the highly sensitive HEMs with less sensitive HEMs. The preparation of new less sensitive HEMs has certain limitations such as chemical compatibility, poor stability, and decomposition performance [3]. Hence, more research is devoted to improving the performance of the existing less sensitive HEMs. 3-nitro-2,4-dihydro-3H-1,2,4-triazol-5-one (NTO)'s highly insensitive, ease of manufacturing, safe handling, and high detonation velocity make NTO one of the most promising candidates [4,5]. The improvement in the thermal decomposition of NTO can be achieved by (1) the use of additives [3,4,6], (2) reducing the size of NTO [7], and (3) formation of co-crystal with other HEMs [8,9]. One of the most widely studied additives for HEMs is transition metals and their analogs [3,10,11]. The transition metal oxides can act as an effective additive due to (1) availability of a wide range of oxidation states that can participate during the thermal decomposition process, (2) availability of large surface, and (3) activeness to participate in the reactions. The transition metals of the 3d series specifically Fe, Co, Ni, Cu, and Zn are known to possess good catalytic activity for reducing the thermal decomposition temperature of HEMs. Transition metal ferrites are widely employed in catalytic applications because of their effectiveness, ease of synthesis, utilization of less expensive chemicals, and environmental friendliness as they can be easily separated using a magnet due to their magnetic nature [12][13][14]. Prabhakaran et al. [4] have reported that the presence of additives can affect (increase or decrease) the decomposition temperature of NTO.
Cobalt ferrite (CoFe 2 O 4 or CoF) is well known to influence the thermal decomposition of HEM like AP [15]. CoF can be synthesized using various methods such as co-precipitation, sol-gel, hydrothermal, green synthesis, flash auto-combustion, etc. [15][16][17][18]. Although the co-precipitation process produces nanosize particles with a wide range, it has benefits over other approaches because it is simpler to use, less expensive, uses less toxic chemicals, and uses less energy-consuming. Hydrothermal, sol-gel, ultrasound, or microwave-aided methods, in situ self-assembly, thermal breakdown, and chemical coprecipitation methods can all be used to fabricate nanomaterials [19]. For this purpose, various solvents can be utilized. However, the results may differ depending on the solvent, i.e., recrystallization of ammonium dinitramide (oxidizer in propellants) in solvents like ethanol, acetone, propanol, isopropanol/NaF, and dimethyl sulfoxide can alter the morphology of ammonium dintramide [20]. Since NTO is a HEM, it explodes when heated, restricting the use of processes like hydrothermal, sol-gel, or thermal decomposition. NTO contains acidic 'H,' and hence, it can form metal complexes. Additionally, modifications to NTO's morphology may modify the temperature at which it decomposes. As a result, the wet chemicals techniques like chemical precipitation, ultrasound or microwave-assisted solution mixing may yield NTO with altered morphology or NTO metal complex. Additionally, the use of solvent raises the total cost of the NTO: CoF synthesis; therefore, NTO: CoF composition synthesized using wet chemical techniques and ultrasonic or microwave aided solution mixing will not be appropriate to compare the additive impact of CoF. As a result, in the current study, the composition of NTO including nanoadditive was made utilizing straightforward mixing with a mortar and pestle.
This paper reports the effect of CoF on NTO, and NTO with reduced size (r-NTO). CoF was synthesized via the co-precipitation route and characterized using X-ray diffraction (XRD), atomic force microscopy (AFM), scanning electron microscopy (SEM), Raman, and ultraviolet-visible (UV-Vis) analysis. The effect of 5% by mass CoF additive on both NTO and r-NTO was investigated using simultaneous thermogravimetry-differential scanning calorimetryderivative thermal analysis (TG-DSC-DTA). Three isoconversional methods namely Flynn-Wall-Ozawa (FWO), Kissinger-Akahira-Sunose (KAS), and Starink method were used to evaluate the kinetic parameter.

Synthesis of NTO
NTO was synthesized from previously reported methods [21][22][23]. Equal amount of formic acid and SC. HCl was mixed with continuous stirring and refluxed at 102°C. After the completion of the reaction excess formic acid was removed by distillation and 2,4-dihydro-3H-1,2,5-Triazol-3-one(TO) was obtained (MP = 232°C). Thereafter, TO was added to an excess of nitric acid maintaining the temperature below 5°C. The solution was refluxed in a microwave reactor (Uwave-1000; serial no.: UW162) at 60°C, 200 W for 15 min. The reaction progress was monitored using TLC employing Ethyl acetate: n-Hexane (8:2) as a mobile phase. After the completion of the reaction, the solution mixture was poured onto ice, NTO was filtered re-crystallized using ethanol solvent, and dried in an oven overnight at 60°C. (Yield of NTO = 68.5%).

Reduction in size
The size of the synthesized NTO was reduced using the solvent-antisolvent technique [24]. A saturated solution of NTO was prepared in 100 mL THF solvent and sonicated for half an hour. This saturated solution of NTO was added to the chilled n-hexane (\ 0°C, 400 mL) via a syringe under magnetic stirring to obtain small size NTO precipitates (r-NTO). The obtained precipitates were filtered and dried in an oven overnight at 60°C (Yield = 66.17%) (Fig. 1).

Synthesis of cobalt ferrite
CoF was synthesized using the co-precipitation method. Equal quantity of salts of Co(NO 3 ) 2 (0.2 M, 150 mL) and Fe(NO 3 ) 3 (0.4 M, 150 mL) were mixed with continuous stirring (mole ratio of Fe 3? :Co 2? was 2:1) at 30°C. To this mixture, 2 M NaOH solution was added to obtain black precipitation. The precipitation was completed at * 11-12 pH (consumed 2 M NaOH = 125 mL). The obtained precipitates were filtered, washed with warm water, and calcined at 500°C for 5 h. The obtained CoF powder was utilized for the characterization and as an additive.

Incorporation of nanoadditive
For the preparation of NTO ? CoF, NTO and CoF were mechanically mixed using a mortar pestle in a 95:5 mass ratio. Similarly, for r-NTO ? CoF, r-NTO and CoF were mixed in a 95:5 ratio. The prepared compositions were further used for simultaneous TG-DSC-DTA analysis for the evaluation of the thermal decomposition parameters.

Characterization
The UV-Vis analysis was carried out using Shimadzu UV-1800 UV/VIS Spectrophotometer. UV-Vis spectrum was recorded between 200 and 800 nm range in dilute hydrochloric acid solution. Raman analysis was performed by JobinYvon Horiba LabRam, HR800 laser sources with 532 nm wavelengths. Powder XRD patterns were analyzed by a Rigaku Ultima IV powder X-ray diffractometer instrument equipped with a Cu Ka radiation source (k = 0.15406 nm). AFM (AFM, NT-MDT, Ntegra Aura) and SEM (ESEM XL-30 Philips; Netherlands) was employed for the topographic investigations.

Catalytic activity
The catalytic activity of the synthesized CoF particles on both NTO and r-NTO was investigated using simultaneous TG-DSC-DTA (Perkin Elmer, STA 8000) at three heating rates (5, 10, and 15°C/min) from room temperature to 400°C temperature under N 2 atmosphere in a platinum pan. The thermal data was further utilized to calculate the activation energy of NTO and r-NTO in the presence of CoF additive.
As per the ICTAC kinetics committee recommendations [25], three heating rates (5, 10, 15°C/min) were used to evaluate the kinetic parameters. DSC Fig. 1 Scheme for the synthesis of NTO and r-NTO peak area was used to evaluate the extent of conversion (a) at various heating rates using Eq. 1. Where dH/dt represents heat flow. Three isoconversional integral methods FWO (Eq. 2) [26,27], KAS (Eq. 3) [28,29], and Starink (Eq. 4) [30] were used to evaluate activation energy (E a ), and pre-exponential factor (A) at the varying extent of conversion. Where, b, R, T, and g(a) represent heating rate (K/min or°C/ min), universal gas constant (8.314 J/K mol), Temperature (K), and integral reaction model, respectively. Among the 41 reaction models g(a) reported elsewhere [31], the most suitable model for NTO, r-NTO, NTO ? CoF, and r-NTO ? CoF is represented in Table S1. E a was evaluated from the slope and ln A from the intercept of the Plots of ln b, ln (b/T 2 ), and  [33]. The Scherrer formula (Eq. 5) was used to calculate the crystalline size of CoF, and NTO from maximum peak intensity (i.e., CoF = 35.60°(311); r-NTO = 27.24°) of the XRD pattern. The smallest particle size is a crystallite, which usually is a single crystal in powder form.
where k represents wavelength, b represents the fullwidth half maxima (FWHM) of the peak, h represents the peak position, and D is the calculated crystalline size (nm). CoF and r-NTO have predicted crystal sizes of 22.14 and 38.39 nm, respectively. Both estimated and measured sizes support the materials' nanoscale crystalline dimensions. The SEM image of CoF ( Fig. 3a) suggest that particles were having a cubic shape having different sizes. Agglomeration was also observed, which could have been because of the cluster of smaller size particles. The agglomeration give image that suggest a micron size particle. Therefore, AFM image of CoF was used to confirm the size. The formation of nanosized CoF particles was confirmed by the AFM picture of CoF (Fig. 3d). The SEM image of NTO (Fig. 3b) and r-NTO (Fig. 3c) confirms the reduction in size of NTO. The SEM image suggests that morphology of NTO remains the same upon reducing the size (i.e., needle shape crystals). CoF particles display a diverse particle size distribution, as shown in the Figure. The bulk of the particles are found between 69 and 92 nm in size. The size of a crystal is finer than the particle size, and therefore, the crystalline size of CoF was smaller than its particle size. There are several crystals with the same alignment inside the particle, it is instead made up of two or more distinct crystallites. Using Eqs. (6- ) and the XRD data, the structural characteristics of CoF were calculated.
where a, d, e, q, N A , V, and M represent lattice constant, dislocation density, microstrain, X-ray density, Avogadro constant (6.022 9 10 23 molecules/mol), cell volume (V = a 3 ), and molar mass (204 g/mol), respectively. The q is the theoretical density predicted by the lattice constants and the number of atoms in the unit cell. That is the density of the material in the absence of any lattice fault/defect, which does not exist in reality but only theoretically. The values obtained for CoF are reported in Table 1 with previous reported values. The parameters' stated values in Table 1 agreed with those that had been previously published [34]. According to the literature, increasing crystalline size is associated with a reduction in microstrain, which is consistent with the findings of the current study. The differences in the microstrain and lattice parameters show lattice distortions that can be caused by the nanocrystallites' grain boundaries. The associated increase in crystallite size can indeed be ascribed to the lattice's expansion as a result of defects around the atoms caused by the mismatch between ionic radius, which would increase the crystallite size [34][35][36].
The UV-Vis spectrum of CoF, NTO, r-NTO, NTO ? CoF, and r-NTO ? CoF is given in Fig. 4. Every sample demonstrates the typical absorbance in the 200-400 nm wavelength range. The nitration of TO into NTO was confirmed by the absorbance in the wavelength range of 250-400 nm since TO does not exhibit absorption in this range (Fig. S1). Because r-NTO absorbs light at a lower wavelength than NTO does, this indicates that the two compounds are different in sizes. According to the previous studies [37,38]. The same substance exhibits increased absorbance over a wider range of wavelengths when it is larger in size. Because CoF also absorbs in this wavelength range, the absorbance of NTO and r-NTO was higher in the presence of CoF additive than it was without it. Using Eq. 10, the Tauc's plot (Fig. 4inset) was utilized to determine the band gap energy of CoF. Where hv is the photon energy (eV), E g is the band gap energy, a is the absorption coefficient, and B is constant. A Tauc's Figure revealed an estimated band gap energy for CoF of 3.19 eV, which was higher than the previously published value of 2.89 eV  [39]. The optical characteristics are greatly influenced by dependent structural features. In various particle size regimes, electron confinement, Coulomb interactions, and binding energy affect the band-gap energy of the material. It was reported that the optical band gap energy increases with increasing size [40]. This was most likely due to the increase in crystallite size and the decrease in defect states. Another cause for the band gap energy to fluctuate might be an alteration in the extent of hybridization between the Co, O, and the Fe. In addition to that, the impurity phases have had a significant impact on the change of the band gap energy. Co substitution can lead to the formation of certain crystal defects, resulting in confined states in the band structures, lowering the E g [35,41,42].
The change in polarizability for nuclear displacement must be non-zero for a substance to be Raman active. The Raman spectrum of CoF, NTO, r-NTO are given in Fig. 5. CoF gives five Raman active modes in the region of 200-800 cm -1 . The peaks at Raman shift 206, 314, 394, 479, and 687 cm -1 correspond to T 2g , E g , T 2g , T 2g , and A 1g respectively, and correspond to vibrational modes at octahedral and tetrahedral sites [39,43]. The presence of these peaks confirms the formation of spinel CoF. In CoF, the local site symmetry of the cubic lattice was lowered by microstrain. Due to the lack of long-range order, the line broadening is a typical nanoscale characteristic (i.e., particles having less size). The Raman spectrum confirms the formation of NTO as the peaks match that of the reported values [44]. The details regarding the peak position of NTO can be found in Table S2. Upon reducing the size of NTO, (1) the corresponding Raman peaks of NTO are broadened, (2) Shifting of Raman peak position to higher wavenumbers, and (3) increased Raman intensity of the peaks (Fig. 6). It was reported that particles with smaller sizes (e.g.,  nanosize) broaden the peak compared to the corresponding bulk material [45]. The Raman peak's widening indicates that the size of the NTO has successfully been reduced. By calculating the bulk densities of NTO and r-NTO, the same was confirmed. NTO and r-NTO powder had bulk densities of 0.51 and 0.29 g/mL, respectively. The fact that the bulk density of r-NTO powder has decreased indicates that the synthesized r-NTO was indeed smaller in size than NTO since the same amount of r-NTO has greater volume than NTO. It has also been confirmed that r-NTO was smaller in size than NTO by the shifting of the Raman peaks in the case of r-NTO to a higher Raman shift. Raman shift is size-dependent, therefore as particle size increased, the peak shifted to lower Raman shift values, which was often accompanied by a drop in dislocation density and strain [46][47][48].

Thermal decomposition
TG and DTG curves of NTO and r-NTO in the present as well as the absence of CoF are depicted in Fig. 7. Molecular dynamics simulations suggest that more than 10,000 chemical reactions are found taking place during the breakdown of NTO, and the number of reactions rises as the temperature rises. The majority of these chemical reactions are discovered once the quantity of NTO starts to go toward zero, demonstrating that the majority of reactions involve unstable intermediate products [49].  Fig. 8. The effect of various heating rates on DSC curve is given in Fig. 9 and the corresponding DSC data is reported in Table 2. The DSC, and DTA curves are similar. In NTO, no endothermic melting peak was obtained before the decomposition indicating the decomposition of NTO occurs before melting [50]. This fact was in agreement with previous report [7]. The single exothermic peak confirms that the decomposition of NTO occurs in a single step. Both the initial decomposition temperature and peak decomposition temperature of NTO are important. The peak temperature (T m ) along with onset (T o ) and ending temperature (T e ) of NTO decomposition are decreased in the case of r-NTO, and NTO ? CoF. This indicates that both incorporation of CoF, as well a reduction in size of NTO, are effective for decreasing the decomposition temperature of NTO. Initial decomposition of NTO involves C 2 HO 3 N 4 and C 2 H 3 O 3 N 4 as major intermediate products. Hence, It is plausible that CoF can lower the decomposition of NTO by facilitating the formation of these two intermediates. The peak temperature plays an important role as it indicates at that temperature maximum gaseous are released with large heat released [49]. The T m of NTO was decreased by 4.71 and 14.97°C using additive and reduction size approaches, respectively. The total time required for the decomposition was also reduced by 30.78 s for NTO ? CoF, and 39.6 s for n-NTO compare to NTO. The DSC-DTA results indicated that reducing the size of NTO was a better approach for reducing the thermal decomposition temperature of NTO than the use of CoF additive. Further, CoF reduces the onset and ending temperature of r-NTO, but the decomposition time along with T m of DSC peak was increased by 20.58 s, and 0.75°C, respectively. The effect of heating rate shows that DSC curves are shifted to higher temperature values as the heating rate was increased (Fig. 9). This could be because, at a lower heating rate, the sample was heated slowly and spend a long time in the contact with the sample pans leading to decomposition at a lower temperature compared to higher heating rates.
The heat released during the decomposition process is another important factor in deciding HEMs efficiency. The order of heat of decomposition released was (r-NTO and (NTO [ r-NTO ? CoF [ r-NTO [ NTO ? CoF) at a heating rate 5, 10, and 15°C, respectively. No certain trend was observed in the case of the heat of decomposition. Therefore, Fig. 7 TG and DTG curves of NTO, and r-NTO in the presence of a 5% CoF additive judging the value at only a single heating rate will not be accurate. The effect of heating rate on the peak temperature (°C), and heat of decomposition (J/g) is presented in Fig. 10 for a better understanding. As represented in Fig. 10a that in the presence of CoF, DSC peak temperature was reduced as compared to pure NTO, and r-NTO at two heating rate values. Further, the DSC peak temperature of r-NTO was less compared to NTO at all heating rates, indicating reliable results. In the case of the heat of decomposition (Fig. 10b), it was evident that although CoF, and reducing the size of NTO are efficient ways to reduce the decomposition temperature of NTO, the decreased heat of decomposition was a downside of using CoF as an additive and reducing the size of NTO. To overcome the drawback of heat released during the decomposition process, CoF can be added to NTO, as r-NTO ? CoF decomposes at lower temperature with slightly increased heat of decomposition compared to r-NTO. To further evaluate the effect of the additive and r-NTO, kinetics parameters such as activation energy and the pre-exponential factors were evaluated at three heating rates.

Comparison with previous studies
The effect of various additives reported for NTO is presented in Table 3. It was reported that the addition of 1% by mass nanosize ZnO, and NiO can reduce the  DTG peak decomposition temperature of NTO from 270 to 234 and 246°C, respectively. The results were good compared to other additives reported in Table 3, but no data regarding onset temperature was reported. In the present work, the system nNTO ? CoF decreases both the initial decomposition temperature and peak temperature of the DTG curve. Despite the lower content of the additive, the nNTO ? CoF shows a better performance for reducing the decomposition temperature of NTO, than the corresponding system containing other energetic catalysts of Co (i.e., GO-T-Co-T and T-Co; where T = Triaminoguanidine and GO = Graphene oxide). However, the effect of 5% by mass CoF on pure NTO was lower than the GO-T-Co-T and T-Co, this is plausible as the latter catalysts contain high energetic additive (Triaminoguanidine) in addition to large additive content. The utilization of Triaminoguanidine pose safety risks and hence, it should be better to use less hazardous additives such as Cobalt ferrite.
To Conclude, the system rNTO ? CoF, can be utilized as a substitution for NTO as its' decomposition will take place at a lower temperature compared to pure NTO. The additive CoF can also be retrieved    from the NTO using a magnet owing to its ferromagnetic nature.

Kinetic study
The conversion plots (extent of conversion vs. temperature) of NTO, NTO ? CoF, r-NTO, and r-NTO ? CoF at 5, 10, and 15°C/min are presented in Fig. 11. At a lower heating rate, the decomposition process begins and ends at a lower temperature than at higher heating rates. The trend was the same for all samples, i.e., for high heating rates, the conversion process was completed at higher temperature values and vice versa. Further, from the conversion plots, it was evident that in the presence of CoF additive, the conversion begins was completed within a short temperature range as compared to pure NTO, or r-NTO. The temperature at the various extent of conversion at an interval of 0.025 was evaluated at three heating rates for calculation of activation energy (E a ) and the pre-exponential factor (A). The plots of three methods (FWO, KAS, and Starink) for NTO, NTO ? CoF, r-NTO, and r-NTO ? CoF are presented in Figs. S3 and S4, in the supplementary file. The average value of the E a , and ln A are reported in Table 4. The variation in the E a , and ln A with the extent of conversion (a) is depicted in Fig. 12. As visible, the pattern of the changes in the E a and ln A concerning a was not altered in the presence of CoF additive (i.e., NTO, and NTO ? CoF exhibit similar patterns as well as r-NTO and r-NTO ? CoF also exhibit similar pattern). Further, the patterns in the case of NTO, and r-NTO are different. The possible cause for this could be a change in the mechanism pathways. The presence of CoF increases the activation energy along with the pre-exponential factor for the decomposition of both NTO, and r-NTO (Fig. 12). Further, the value of E a and ln A was lower in the case of r-NTO compared to NTO up to halfway through the reaction, (a = 0.5), After that the value of E a and ln A drops drastically. From the activation energy calculations, it was clear that the decomposition of NTO requires high activation energy barrier to surpass (-369.30 kJ/mol) which matches with the reported value of (322-368 kJ/mol) [52], efforts are devoted to decreasing this large activation energy barrier [3].

Conclusions
CoF was synthesized using the co-precipitation method. XRD confirms the crystalline size of CoF falls in the nanoscale region. The CoF show wide particle size distribution with cubic shape. In the presence of CoF, the exothermic curve of NTO was shifted to a lower temperature than pure NTO. NTO ? CoF has 4.71°C lower peak temperature than NTO. r-NTO and r-NTO ? CoF also decomposes at lower temperature than NTO. The lowered decomposition temperature can provide better burning rate performance of corresponding energetic formulations than NTO. However, the activation energy of NTO was increased and heat released was decreased for all three compositions (i.e., r-NTO, NTO ? CoF, and r-NTO ? CoF). These limits the NTO ? CoF's use as an energetic formulation.