Figure shows three SEM images of the mixed Al nanoparticle and NiO nanowire composite before (Figure a) and after (Figure b,c) sonication. Figure a demonstrates the sizes of Al nanoparticles (about 80 nm) and the diameter (about 20 nm) and length (about 1.5 μm) of NiO nanowires after mixing two components. These distinct images of two components show a poor dispersion of nanoparticles in the network of nanowires. After the solution was sonicated and dried, Al nanoparticles were able to decorate on the NiO nanowires, as shown in Figure b. A higher-resolution SEM image shown in Figure c demonstrates the nanowire branches beneath the Al nanoparticles. This process was expected to significantly increase the contact area between two components, improving thermite performance.
SEM images of Al nanoparticle and NiO nanowire composites before (a) and after (b, c) sonication. Scale bar 100 nm in (a), 2 μm in (b), and 100 nm in (c).
Figure shows several DSC/TGA thermal analysis curves measured from three Al nanoparticle and NiO nanowire composites with different NiO weight ratios (for samples B, D, and E, respectively, in Table ). Note that the heat flow curves in Figure a,b,c were plotted using the mass corrected values. Figure a was measured from sample B which originally contained about 4.2 mg of material and with a NiO weight ratio of 20%. When the sample was heated from room temperature, a slow mass loss was observed at a low temperature range (<390°C) which was attributed to the dehydration of the sample. When the temperature was increased above 400°C, the mass of the sample first increased then decreased. This behavior was associated with the mass change before and after the thermite reaction (in comparison with the heat flow curve). When the temperature is close to the onset temperature, the Al core inside the Al nanoparticles exposes to the surrounding through diffusion through or breaking the Al2O3 shell. The Al element can react with the surrounding gas such as water and oxygen if the purging flow rate is insufficient, which causes the mass increase around the ignition temperature. An obvious mass loss was observed from 450°C to 550°C, which was attributed to the formation of gas and vapor from the thermite reaction. As mentioned earlier, these gas-phase species can include vapors of metals (such as Ni) and the elemental oxygen. In this figure, the mass loss process stopped when the maximum heat flow was generated from the sample (which indicates the most energy available for vaporizing the metal products). Following the thermite reaction, the mass of the sample increased almost linearly. On the accompanying heat flow curve, the energy generation from the thermite reaction is clearly visible between 450°C and 550°C. The onset temperature was measured as 450.1°C from this curve. The area integration based on this heat flow curve provided the energy release per unit mass of the composite of about 321 J/g.
DSC and TGA profiles measured from these Al/NiO MIC with different NiO ratios. (a) Sample B 20 wt.% NiO, (b) sample D 33 wt.% NiO, and (c) sample E 38 wt.% NiO.
Figure b shows the measured data from sample D which contained about 2.8 mg of material and with the NiO weight ratio of 33%. A multistage mass loss process was observed in the low-temperature range between room temperature and 475°C, due to hydration, and the possible decomposition of NiO. Note that for this measurement, there was little mass gain observed before the ignition of the thermite reaction, which indicates a sufficient purge process, as discussed before. A sharp mass loss was observed when the thermite reaction occurred. Again, this mass loss process stopped when the maximum heat flow was generated from the sample. On its heat flow curve, the thermite reaction was observed between 480°C and 550°C. The onset temperature for this exothermic peak was measured as 484.0°C. The energy release per mass value was determined as 593 J/g for sample D. Note that sample D produce more energy per mass due to the increased NiO amount in the composite. Figure c was measured from sample E which contained about 3.6 mg of material and with NiO weight ratio of 38%. The mass change and heat flow curves are very similar to these data taken for sample D. The onset temperature was measured as 475.0°C. The energy release per mass was calculated as 645 J/g. Note that the energy release values were measured by accounting for the total mass of the Al nanoparticles and NiO nanowires. Since the Al content was assumed as 42% in these Al nanoparticles, the following equation was used to determine the energy release per unit mass of the pure Al and NiO composite:
where E (J/g) is the energy release per mass of MIC, E′ (J/g) is the DSC curve-determined energy release per mass, mAl,Al2O3,NiO (mg) is the total mass of the composite, and mAl,NiO (mg) is the mass of the total Al content in Al nanoparticles and NiO nanowires. Because the DSC measurements were conducted in a non-adiabatic condition, the values of E are much smaller than the theoretic reaction enthalpy of the reaction R2. Figure shows the calculated E values and the measured onset temperatures from the samples shown in Table .
DSC-determined onset temperatures and energy release values for Al/NiO MIC with different NiO ratios.
The dependence of the onset temperatures on the NiO ratios of the composites is shown in Figure . It can be observed that increasing the NiO ratio did not significantly change the onset temperature of the exothermic peak. This indicates a narrow size distribution of Al nanoparticles in these composites and sufficient intermixing between Al nanoparticles and NiO nanowires. All measured onset temperatures are smaller than the melting temperature of bulk Al. In the literature, it was suggested that the activation energy of the thermite reaction depends on the diffusion distance over which these metal ions (aluminum and nickel which become available from the decomposition of NiO) need to travel before initiating the reaction [46
To quantify the activation energy of the Al nanoparticle and NiO nanowire composites, the DSC curves of sample D was processed directly using the TA software and through the implementation of the American Society for Testing and Materials E698 method. Note that the ASTM method is often the only effective approach to analyze reactions with multiple exotherms because these peak temperatures at different heating rates are not significantly influenced by the baseline shift [47
]. The ASTM E698 method generally gives an accurate assessment of the activation energy. However, calculations of the pre-exponential factor (Z
) assume the n
th order reaction behavior. The derived activation energies for sample D are 216.3 and 214.5 kJ/mol, respectively, from two methods. Figure shows the procedure to determine the activation energy from the DSC data when the kinetic rate was expressed as a function β
) of the temperatures Tmax
corresponding to the maximum heat flow. The derived activation energy agrees generally with the previously reported activation energies for Al nanoparticle-based thermite composites (such as, 248, 222, and 205 kJ/mol for the Al-Fe2
, and Al-MnO3
, respectively [48
]). The activation energy of the Al nanoparticle and NiO nanowire MIC is close to but lower than the reported activation energy of the NiO reduction process (277 KJ/mol [49
]). Taking into account the size effect on the reactivity of NiO nanowires, this ignition energy may indicate a thermal decomposition of NiO about the onset temperature of the studied MIC, which behaves similarly to the ignition of the Al-Bi2
]. Meanwhile, for heterogeneous condensed phase MICs, the limiting factor affecting the ignition event can also be the solid-phase diffusion. Further investigations on the ignition mechanism of the Al/NiO MIC are expected.
Graph used for determining the activation energy of sample D, 33 wt.% NiO, using ASTM E698 method.
The XRD analysis was performed on the reaction products from sample D which was a fuel-rich MIC with Φ
= 3.5. As shown in Figure , the chemical compositions were identified as Ni, Al, AlNi, and Al2
. Note that the identification of Al2
using XRD is evidential from the previous study [51
]. In addition to those solid products, gaseous species such as O2
was also possibly formed. It is interesting to reveal the production of AlNi from the Al/NiO MIC. As a comparison, the formation of Ni was shown with lower and fewer XRD peaks, while Al still existed as a relatively large amount. Based on these observations, the following reaction was responsible:
XRD patterns measured from the reaction product of sample D, 33 wt.% NiO.
Note that in this study, MIC Φ = 3.5 contained abundant Al nanoparticles and thus made the reaction R7 feasible. The propagation of R7 does not necessarily require the completeness of R2 since the decomposition of NiO may occur first and be followed by the reaction between Al and Ni. A further study on elementary reactions related to R2 and R7 is needed in order to gain more insights on this issue. To further characterize these microstructures of the products, the SEM and EDAX analyses were performed on the same product examined by XRD. Figure shows two typical structures observed from MIC Φ = 3.5: (Figure a,c) a sphere which was rich in Ni and Al, and (Figure b,d) a bunch of Al2O3 crystalline structures. The coexistence of Ni and Al in the sphere is possibly in the form of AlNi.
SEM images (a, b) and respective EDX patterns (c, d). They were obtained from the reaction products of sample D, 33 wt.% NiO.
In order to further examine the possible formation pathway of the AlNi phase, ab initio MD simulation was conducted for scoping the reaction time scale and identifying the equilibrium product of the thermite reaction of the Al/NiO MIC. For this simulation, the initial temperature was set to 0 K. At this temperature, the thermodynamic equilibrium structure of an Al crystalline nanoparticle and a NiO nanowire was obtained, as shown in Figure a. The system temperature was then increased to 1,000 K (or 726°C) to ignite the reaction. After ignition, the simulation was done under adiabatic condition. It was found that after 5 ps, as shown in Figure b, Al atoms diffused through the Al-NiO interface and met with O atoms (while the diffusion of O atoms into the Al nanoparticle was possible but with a much smaller chance, as observed from the image where only one O atom was found in the Al nanoparticle). Meanwhile, Ni atoms were grouped together and were intended to form the pure Ni phase. It was also observed that the AlNi phase exists at the interface between the Al nanoparticle and the NiO nanowire. Accompanying this fast thermite process, the system temperature was increased up to 3,500 K within 5 ps. This MD simulation confirmed the possibility of forming the AlNi phase from the Al-NiO thermite reaction and revealed the diffusion paths of Al and Ni atoms during the thermite reaction.
Two MD-simulated equilibrium structures of the adjacent Al nanoparticle and a NiO nanowire. (a) Temperature = 0 K; (b) temperature = 3,500 K.