2.1 Morphological Analysis of Al/CuO/V2C
The Al/CuO/V2C samples prepared with different concentrations of V2C MXene suspension result in diverse morphologies. As illustrated in Fig. 1a, although from the same procedures, the samples with lower concentrations (≤ 5 wt%) show a planar structure, while the ones with higher concentrations (> 5 wt%) tend to have a spherical structure. It turns out that the CuO and Al nanoparticles go through a self-assembly process with V2C nanosheets in the suspension. However, in an ordered sequence rather than a random combination. When both CuO and Al particles are present, CuO has a higher tendency to bind on the V2C surface. The spontaneous self-assembly procedures begin with the formation of CuO/V2C precursor, and then the resulting Al/CuO/V2C nanocomposite in both low and high concentration cases. This unique phenomenon suggests different types of bonding occur between Al, V2C, and CuO, V2C in terms of bonding strength and distance. Specifically, the bond strength between CuO and V2C surface is higher than that of Al and V2C, whereas the bond distance is shorter in the former case. From the SEM image of the low-concentration (1 wt%) Al/CuO/V2C sample (Fig. 1b), the original morphology of the V2C nanosheet is preserved after the assembly, with Al and CuO nanoparticles covering the surface. In the TEM image (Fig. 1c), following ultrasonic treatment of Al/CuO/V2C, the CuO/V2C precursor film reveals a dense attachment of plate-like CuO particles onto the V2C surface. Through the EDS analysis on the HAADF-STEM images (Fig. 1d-g, Figure S1), both O, V, Al, and Cu are evenly distributed amongst the composite, in which the distribution of Cu, and O are almost identical, confirming the presence of CuO. Besides, although the Al element is also detected, from the low weight fraction (< 1 wt%) and similar distribution to the V element, it might be due to the residue Al inside V2C due to incomplete etching of MAX phase. The high-resolution TEM image (Figure S2) evidences the adhesion of CuO on the V2C surface. The interplanar spacing of CuO and V2C is 0.24 nm for the (1–11) plane and 0.25 nm for the (0-110) plane respectively25, and the contact line is marked with a red dashed line.
The general morphology of the Al/CuO/V2C composite changes drastically as the concentration of V2C increases. In Fig. 1h, microspheres comprised of nano-sized particles are observed. The TEM images (Fig. 1i-m) further unveil the internal structure. It appears that the V2C nanosheet curls its surface, while the CuO and Al nanoparticles are present both in and out of the envelope. The mapping results prove the existence of both Al (Fig. 1l) and CuO (Fig. 1j, m, Figure S3) particles, indicating the successful incorporation. This also suggests the difference between the binding characteristics of Al and CuO. Compared with the V2C distribution (Fig. 1k), CuO shows more overlapping area than Al and, therefore more affinity to the V2C surface.
2.2 Compositional and Elemental Analysis of Al/CuO/V2C
To further verify the composition of obtained samples, PXRD and XPS analyses have been carried out. From the PXRD result (Fig. 2a), the as-synthesized Al/CuO/V2C contains Al, CuO, and V2C, as it shows both characteristic diffraction peaks of the 3 components, each of which is marked with specific symbols. The peaks at 8° and 12° represent single26 and multiple27 layers of V2C respectively. However, The V2C diffraction peaks are hardly detectable in the sample, which is mainly due to low peak intensity, small quantity, and possible overlap with Al and CuO peaks. From the XPS survey scan (Fig. 2b), the major elements of Al/CuO/V2C such as Al, Cu, O, V, C, and a small amount of F, are identified. The deconvoluted V 2p/O 1s region (Fig. 2c) shows peaks representing C = O, C-O, and Metal-O bonds, respectively28. Where V5+ is present in V 2p orbital, indicating the surface oxidation of V2C after the assembly. In the Cu 2p region (Fig. 2d), Cu2+ is presented with 2 satellite peaks, suggesting that CuO is not reduced throughout the preparation process.
2.3 Thermal Analysis of Al/CuO/V2C
The thermal properties of energetic materials are central to their overall performance and safety. In this regard, a DSC-TG analysis has been conducted. Firstly, the effect of V2C amount is discussed. In the air atmosphere, Al/CuO nanocomposite releases heat at 500–850 ℃ with 2 separated peaks (Fig. 3a). The addition of V2C gradually postponed the onset temperature of the first peak from 510 ℃ (1 wt%) to 560 ℃ (5 wt%) and stabilized at higher concentrations. Meanwhile, the onset temperature of the second peak remains after the addition. The postponement is largely attributed to the high thermal conductivity of V2C that effectively absorbs the heat from neighboring Al and CuO particles. Besides, the peak temperatures of Al/CuO/V2C are also elevated from 580 ℃ (1 wt%) to 620 ℃ (5 wt%), which suggests the thermite reaction mechanism shifts from Al/CuO to Al/CuO/V2O5 as the concentration increases29. The second peak is more complicated as the concentration increases, in which two exothermal peaks at 770 ℃ and 790 ℃ are predominant, each representing a thermite reaction of Al/CuO and Al/V2O5. According to the integral of DSC curves, the heat of the reaction (Fig. 3b) increases along with the concentration of V2C from 1300 J/g (1 wt%) to 3156.2 J/g (10 wt%), then decreases to 1600 J/g (20 wt%). The addition of V2C has a sevenfold improvement maximum on the heat release of Al/CuO thermite. Besides, the second exothermic peak contributes more to the total heat release, from 45% (1 wt%) to 70% (10 wt%). Although the heat from the oxidation of Al (10870.4 J/g, Figure S4) and V2C (6471.3 J/g, Figure S5) by air contribute to the total heat of reaction, the actual heat release exceeds the sum of each component. This suggests the catalytic effect of V2C. From the TG curves (Fig. 3c), the thermite reaction of Al/CuO/V2C in air atmosphere is divided into 3 stages: in the first stage (200–500 ℃), the main reaction is the loss of the functional groups of V2C, leading to a weight loss that is proportional to the amount of V2C; in the second stage (500–700 ℃), the curves resemble the exothermic peaks in terms of onset temperature, and the weight gains due to the oxidation of Al and V2C are associated with the heat release; in the final stage (700–850 ℃), the weight gain caused by Al and V2C oxidation is also related strongly to the amount of heat release in the corresponding DSC curves.
The samples with different equivalent ratios behave distinctively. For the Al/CuO/V2C samples with 5 wt% V2C, the DSC curves (Fig. 3d) indicate that the heat release is maximum when Al is slightly excessive (Φ = 1.3), while the lack of fuel (Φ = 0.8, 1.0) appears to have more impact on the heat release of Al/CuO/V2C compare with the lack of oxidant (Φ = 1.5, 1.8), which suggests that V2C tend to behave as oxidant rather than fuel. Additionally, the reaction heat in the first exothermal peak (part 1) increases along with Φ, whereas that in the second peak (part 2) shows a maximum at Φ = 1.3 (Figure S6).
To better comprehend the role of V2C in the reaction process, the reaction products (from 10 wt% V2C samples) are taken after heating at different temperatures in the air atmosphere and then characterized using PXRD (Fig. 3e). The results show that at 350 ℃, after the release of surface terminations, a small amount of Cu3V5O4 is detected due to the reaction between CuO and V2C30. At 700 ℃, the peak intensity of Al is sharply decreased, whereas that of CuO remains, also a small amount of Cu2V2O7 is detected31. This indicates the oxidation of Al and V2C as well as the reaction between CuO and V2O5. At 850 ℃, Al2O3 emerges as Al has completely been oxidized, where CuAl2O4 is detected due to a reaction between CuO and Al2O3 at high temperatures32,33. Here the formation of copper vanadate is of particular interest. According to the morphology of the samples, CuO shows a high affinity to the V2C surface with a large contact area, providing chances for the later formation of copper vanadate. Once the vanadate is formed at the interface, oxygen can readily diffuse from V2O5 to CuO through Cu-O-V bonds, changing the mechanism from Al/CuO to Al/CuO/V2O5. From the XPS analysis on the reaction product (Fig. 3f), the intensity of V5+ remains during the reaction process compared with those in the argon atmosphere in different concentrations of V2C (Figure S7), indicating the high reactivity of V2C surface to absorb oxygen from the air and prevent reduction during the reaction process.
The effect of atmosphere is also discussed. The DSC curves from the Al/CuO/V2C samples in the argon atmosphere show two types of thermite reaction from 300 to 800 ℃ as separated by the dashed line (Figure S8). The curves for concentrations lower than 5 wt% have similar behavior to the ones in the air atmosphere, while curves for higher concentrations suggest more subtle reactions with separated exothermic peaks. The onset temperature is also advanced to 400 ℃ at most (10 wt%). The TG curves in the argon atmosphere (Figure S9) also suggest the detachment of surface terminations of V2C before 350 ℃ (stage 1). However, the subsequent weight gain in the stage 2, and 3 is less in correlation with the DSC curves, especially for the higher concentration samples, which is probably due to the continuous participation of oxygen in air. For the 10 wt% samples in the argon atmosphere (Figure S10), the reduction of CuO began as early as 400 ℃ by the presence of Cu2O. Vanadium is detected at 500 ℃ as a sign of the thermal decomposition of V2C. At 600 ℃, the Cu2O is further reduced to Cu. The presence of Al2O3 peaks and disappearance of Al peaks at 700 and 800 ℃ indicate that the Al/CuO thermite reaction has completed.
2.4 Energetic Performance of Al/CuO/V2C
An open burn experiment is carried out to assess the energetic performance of Al/CuO/V2C (Video S1-4). The Al/CuO sample without the addition of V2C burned for 6 ms total, and the flame reached a maximum of 1 ms (Fig. 4a). After adding 1 wt% V2C, the total burning time is reduced by half (3 ms), and the flame is maximum at 0.4 ms with a great enhancement in terms of covering area (Fig. 4b). For the sample with 5 wt% V2C, the burning performance is similar to that of the 1 wt% sample, with a short postpone of burning, and the maximum covering area occurs at 0.5 ms (Fig. 4c). The property of flame changes as the concentration of V2C further increases. In the 10 wt% sample case, instead of an explosion in former samples, it steadily burns until the reaction is complete, creating a flame that lasts for 16 ms (Fig. 4d). The drastic change in the burning property of samples is mainly induced by the evolution in morphology (Fig. 4e). For low concentrations, Al/CuO/V2C forms a planar structure, where Al and CuO nanoparticles attach to the V2C surface successively. In this manner, the agglomeration of nanoparticles is alleviated, bringing more contact area between Al and CuO particles, thus enhancing the average diffusion rate of oxygen, which results in shorter burning time and higher energy release rate. However, for high concentrations, Al/CuO/V2C forms a quasi-spherical structure where V2C is the shell and Al, CuO particles are located both inside and outside of it, resisting the particles inside to explode freely, which reduces the reaction rate significantly from explosion to combustion while largely increase the energy release efficiency and duration.
Lowering the pressure is essential for nanothermite in both safety and microinitiator applications 1. Therefore, the closed-bomb test is conducted to test the combustion property in a confined space as well as the generating ability of Al/CuO/V2C samples. The dynamic pressure of Al/CuO/V2C samples is lower than that of Al/CuO, reaching a maximum of 0.1 s (Fig. 4f). The lowered pressure is mainly due to the introduction of gasless V2C, and partly because of the insufficient oxygen that slightly alters the reaction mechanism. The trend of pressurization rate is identical to the peak pressure for Al/CuO/V2C samples (Fig. 4g). Generally, the peak pressure and pressurization rate decrease as the concentration of V2C increases, whereas it shows dependency on the structure of Al/CuO/V2C (grey region). The general trend is mainly caused by the decrease in gas generation since the decomposition of V2C contributes little to the gaseous product. However, the peak pressure and pressurization rate of Al/CuO/V2C with 5 wt% and 7.5 wt% is similar to those of the 1 wt% sample, indicating the role of structural difference. The peak pressure and pressurization rate of Al/CuO/V2C were reduced by 55.4% and 57.5% respectively (from the comparison of 0 and 10 wt% samples), which means V2C can greatly improve the safety of Al/CuO and is feasible for microinitiator application.
2.5 Binding Mechanism Analysis of Al/CuO/V2C
As the highly tunable performance of Al/CuO/V2C is driven by the unique structures deeply rooted in the ordered self-assembly of the components, the mechanism behind the assembly is worth further exploration. Al/V2C and CuO/V2C samples are fabricated and characterized by PXRD (Figure S11). The possible covalent interaction at the CuO/V2C interface is examined by Raman spectroscopy (Fig. 5a). A characteristic peak at 895 cm− 1 representing the vibration of the Cu-O-V bond is found in the CuO/V2C sample while being absent in other samples, which is direct evidence of the covalent bond at the CuO/V2C interface 23. The absence in the Al/CuO/V2C spectrum is probably because the CuO/V2C interface is more densely covered. From FT-IR spectroscopy (Fig. 5b), the -CH2-CH2- vibration is identified with multiple characteristic peaks in both Al/CuO/V2C and CuO/V2C samples while being absent in other samples. As -CH2-CH2- vibration is commonly present in graphene samples 34, one reasonable deduction is that V atoms at the interface form Cu-O-V bonds by breaking V-C bonds, leaving a small fraction of carbon nanosheet, which indirectly supported the existence of Cu-O-V bonds. From the XPS analysis result (Fig. 5c), both Cu2+ 2p3/2 and Cu2+ 2p1/2 orbits shift slightly to higher binding energy, which also suggests the appearance of the Cu-O-V bond 35. Al nanoparticles exhibit a more randomized distribution around CuO/V2C and a weaker binding strength compared with the covalent bonding between CuO and V2C, therefore the assembly mechanism is presumably the electrostatic attraction. From the zeta potential analysis (Fig. 5d), the mean zeta potential of Al and CuO particles is positive, while that of the V2C suspension is negative. After mixing with a low concentration (1, 5 wt%) of V2C, the zeta potential of the system is more positive than each component, mainly because V2C reduces the agglomeration of the nanoparticles while influencing little to the overall zeta potential. As the concentration increases (10, 20 wt%), zeta potential drops sharply until reaching that of pure V2C, suggesting that the charge is neutralized by V2C.
Based on the binding properties, it is clearer how the layered structure of Al/CuO/V2C from ordered self-assembly facilitates the thermite reaction in terms of oxygen diffusion. As depicted in Fig. 5e, for Al/CuO nanocomposite upon ignition, the oxygen atoms begin to diffuse from CuO to the neighboring Al particles through the Al/CuO interface. However, the random assembly of particles and agglomeration tend to result in incomplete thermite reaction due to poor contact area between Al and CuO. In the Al/CuO/V2C case (Fig. 5f), once ignited, oxygen atoms can not only migrate easily from CuO to nearby Al particles, which benefits from the larger interface area brought by the layered structure but also be replenished from V2C through Cu-O-V bond in air atmosphere, leading to the complete oxidation of Al into Al2O3. To better show the difference between the ordered self-assembly and as-reported electrostatic self-assembly method, the Al/CuO/V/GO control group is fabricated with the same elemental composition to the Al/CuO/V2C sample. From the comparison of DSC curves from Al/CuO/V2C and Al/CuO/V/GO control group (Figure S12), the exothermal peaks of Al/CuO/V/GO are separated into Al/V2O5 (622 ℃) and Al/CuO (567 ℃) peaks. The result indicates that the connection between the components of Al/CuO/V/GO is poor, and part of Al only reacts with CuO while the other part only reacts with V2O5. This finally leads to a lower heat release due to incomplete thermite reaction and agglomeration. Unlike previously published additives, V2C improves both the structure and reactivity of Al/CuO. Besides, the highly tunable energetic performance from its structural evolution enlightens more possibilities for future application.