Reactive sintering: An important component in the combustion of nanocomposite thermites
Introduction
Nanocomposite thermites, or metastable intermolecular composites (MICs) are intimate mixtures of metal/metal oxide nanoparticles, and typically have the consistency of a loose powder. Using nanoparticles greatly reduces mass diffusion lengths between the fuel and oxidizer, and also increases the interfacial contact and homogeneity of mixing. Upon ignition, these materials give rise to a self-propagating reaction with a characteristically high temperature, and low to moderate gas production. Research on MICs can be traced back about 15 years to when Aumann et al. [1] showed that using nanoparticles of Al/MoO3 resulted in several orders of magnitude increase in combustion characteristics over similar mixtures with micron-sized particles. Since then, research efforts have increased to understand the ignition and combustion mechanism, so that improvements in safety and performance can be achieved.
MICs have been experimentally shown to exhibit pressures and flame velocities somewhere in between propellants and explosives [2]. Flame velocities range between 10’s to 1000’s of meters per second, while the pressures range between a few to nearly 1000 atmospheres. The pressure and flame velocity in MICs is something that can be tuned through easily-adjusted parameters, such as changing the method and uniformity of mixing [3], [4], [5], particle size and distribution [6], [7], [8], choice of materials and stoichiometry [9], mixture density [10], or by other techniques such as electrostatic assembly [11] or creating new types of core–shell oxidizers [12]. This tunability, along with other attributes such as high mass/volumetric energy densities and the production of environmentally benign products, make MICs very attractive energetic systems. Nano-sized aluminum is a fuel which exhibits a high reactivity, and nano-Al based MICs are currently being investigated for uses in propellants, explosives, and pyrotechnics, along with other more recent applications, such as antimicrobial energetic systems designed to combat harmful biological agents [13].
Despite the amount of experimental results available in the literature, the ignition and combustion mechanism remains poorly understood. A major problem has been designing experimental techniques which can probe the intrinsic reaction while replicating the environment these materials are subjected to during the self-heating in a freely propagating reaction. This means very rapid and uniform heating, speculated to be somewhere in the range of 4 × 104 K/s (Martirosyan et al. [14]) to upwards of 108 K/s, predicted by an ad hoc calculation assuming thermites can reach an ignition temperature of ∼1000 K in 10 μs, an experimentally observed pressure rise time [15]. Furthermore, in order to understand the thermite mechanism, the ignition and combustion mechanism of nano-Al itself must first be well understood.
Nano-Al forms a passivating oxide shell when exposed to air. This shell is amorphous and uniform [16], and typically has a thickness of 2–3 nm [17]. The oxide shell can occupy a relatively large portion of the particle’s mass, and in some cases can even exceed 50 Wt% [18]. One result generally observed is that the measured ignition temperature of nano-Al is much closer to the melting point of Al, and far below the melting point of Al2O3, which is where large aluminum is reported to ignite [17]. This observation has led to much speculation that the interaction between the low melting point core (933 K) and the high melting point shell (2327 K) is critical to elucidate the underlying mechanism leading to ignition.
Several authors have focused on understanding the ignition mechanism of nano-Al from the standpoint of the core/shell interaction. Trunov et al. [17], [19] used thermal analysis (∼40 K/min) and developed an ignition mechanism suggesting that outwards transport of aluminum can occur as the oxide shell undergoes polymorphic phase transformations, rendering it permeable to aluminum. Using hot stage microscopy at similar heating rates and under vacuum, Rai et al. [20] have observed the outwards flow of aluminum through the oxide shell above the melting temperature. In a separate work, it was shown using molecular dynamics simulations that the effective transport rate of aluminum through the shell can significantly be enhanced by built-in electric fields [21]. At even higher heating rates (∼106 − 108 K/s), Levitas et al. [22], [23], [24], [25] have suggested that a “Melt Dispersion Mechanism” can occur. In such a mechanism, the rapid melting of the aluminum core induces significant stress in the oxide shell to completely rupture it, followed by the subsequent unloading of the aluminum core via tensile stresses. The authors have suggested that in such a mechanism, the kinetics would not be limited by diffusion. A key parameter for this mechanism is the relative thickness of the oxide shell to the aluminum core, and the authors have shown in some cases that increasing the shell thickness can lead to a higher flame velocity, a result which is consistent with such a mechanism.
Understanding the oxidation mechanism of nano-Al is even more challenging when one considers that the mechanism is strongly contingent upon understanding the relevant ignition mechanism. Also, experimental techniques which employ uniform and rapid heating rates are necessary to replicate a true combusting environment. Rai et al. [27] have investigated nano-Al oxidation at intermediate heating rates (∼103 K/s) using a furnace, and have suggested that oxidation is diffusion-controlled via a shrinking core mechanism [28], which involves inwards diffusion of O2 through the oxide shell. At even higher heating rates, (∼106 − 108 K/s), Bazyn et al. have conducted several experiments of nano-Al burning in varying environments inside a shock tube [29]. The authors used optical pyrometry to measure the combustion temperature of the particles as a function of pressure and gas composition, and suggest that the burning cannot be modeled by a droplet burning model, but instead large heat losses characteristic of nanoparticles cause the flame to sit much closer, if not directly on the particle surface. This suggests that heterogeneous reactions between the gas and the particle are prominent in the combustion mechanism. The authors have also investigated the ignition and combustion of nanocomposite Al/Fe2O3 and Al/MoO3 using the same technique, and measured the ignition temperatures in an inert environment to be 1400 and 1800 K, respectively [30]. It should be noted that these ignition temperatures are significantly higher than the melting temperature of Al, which has been experimentally observed to be approximately where nano-Al ignites in a gaseous oxidizing environment [17]. The authors also measure the combustion temperatures of the composites to be in the range of 2750–3350 K, and find that combusting in an oxygenated environment can raise the temperature several hundred degrees, indicating some degree of reaction with the gas.
Besides the aforementioned experiments, there have been limited other studies of the ignition and combustion of nano-Al and nano-Al composite materials which:
- (a)
Avoid the negative effects of studying a bulk sample such as packing density, mixing, differences in heating, etc.
- (b)
Probe intrinsic properties.
- (c)
Uniformly and rapidly heat the samples.
These considerations have led to the development of temperature jump (T-Jump) techniques, which can ramp the temperature of a small amount of sample very quickly. In these experiments, a thin wire or filament (coated with sample) is supplied a tunable voltage pulse and rapidly heats (∼105 − 106 K/s) through resistive heating. The ignition and combustion event can be monitored optically [31], [32], or in a mass spectrometer [33] to probe transient species evolution. Chowdhury et al. [31] used this setup to examine the ignition delay in a nano-Al/CuO composite as a function of aluminum oxide shell thickness. The authors concluded that the diffusion of Al through the oxide shell was responsible for the delay, since an increasing delay time was measured with increasing oxide shell thickness.
One other phenomenon which has received little attention in nanoparticle combustion studies, and is relevant to this work, is the sintering of aggregated particles. This directly impacts the question of size dependence on reactivity, and what is the “effective” particle size of the reacting material. Commercially available nanoparticles are almost always highly aggregated, and the size specified by a supplier often corresponds to the average size of the primary particles within these aggregates. Surface tension forces will of course drive the particles to coalesce if the temperature is sufficiently high to make the particles liquid-like [34], [35]. In a reacting thermite, nanoparticles can be rapidly heated by hot interstitial gases, and can be further heated by the energy liberated during an exothermic reaction. If the exothermic reaction is primarily responsible for inducing the sintering processes, it can be referred to as reactive sintering, a phenomenon which has been shown to be important in Al/Ni reacting systems [36].
One key point we will have to consider is how the kinetic timescale for sintering compares with reactive timescales we observe experimentally. If sintering occurs on a comparable timescale, or faster, than the chemical reaction, it raises two very important questions:
- 1)
Do nanoparticles maintain their high surface area morphology during combustion, and if not, then what is the appropriate “size” to report?
- 2)
Is there an advantage of using aggregated nanoparticles below a certain critical size?
The current work is a compilation of various experiments investigating the reaction mechanism of both nano-Al and nano-Al thermites. The studies are all performed at high heating rates (105 − 106 K/s), by passing a tunable voltage pulse through either a thin Pt wire or a specially fabricated electron microscopy grid. Several different thermite systems were investigated, and not all were studied using every experimental technique, largely due to time constraints on borrowed equipment or facility usage. The particular thermite studied in each case, therefore, was selected based on what would give the clearest representation of the steps involved in the nanocomposite thermite reaction for the particular experimental technique. The experimental results will suggest that a general similarity for all systems studied is that at least some amount of condensed-phase reaction and sintering occurs.
Section snippets
Experimental
The nano-Al used in this work is termed “50 nm ALEX”, and was purchased from the Argonide Corporation. The average primary particle size is specified by the supplier to be 50 nm, and the elemental portion of the particles was found to be 70% by mass, measured using thermo gravimetric analysis (TGA). A representative image of the nano-Al is shown in Fig. 1. The primary particles are largely spherical in nature, and are highly aggregated. One of the samples of CuO, which we term “6nmCuO”, was
T-Jump/PMT ignition temperature
The ignition temperature, is summarized for various thermite systems in Table 3. Also included in the table is the bulk melting point of the oxidizer. While both WO3 and Bi2O3 do melt, the melting of CuO and Fe2O3 involves the decomposition to Cu2O and Fe3O4 (and FeO at higher T), coupled with the release of O2 gas. We have recently argued, through temporally resolved mass spectrometry, that the O2 release for these particular oxidizers plays an important role in the ignition and combustion
Characteristic reaction and sintering times
Up to this point, we have shown visual evidence that sintering is occurring, however, we have only really discussed it in the context of a reactive sintering mechanism. That is, the exothermic reaction initiates, and this causes rapid melting/fusion of adjacent particles. In general, particle sintering is a thermally activated process, and will occur even in the absence of an exothermic reaction, as long as the temperature is sufficiently high. The following section investigates the kinetic
Conclusion
This reaction mechanism of nano-Al based thermites using several high heating techniques was investigated. First, thermites were rapidly heated on an ultra thin Pt wire, and the optical emission was monitored to determine the ignition temperature. It was found that three nano-Al based thermites (CuO, Fe2O3, WO3) ignited above the melting temperature of Al, with Al/Bi2O3 igniting just below this temperature.
High heating rate microscopy experiments were conducted for pure nano-Al and CuO, along
Acknowledgments
We thank the Army Research Office and the Defense Threat Reduction Agency for their financial support. We acknowledge the support of the University of Maryland’s Nanocenter and its NISPLAB. The NISP laboratory is operated jointly by the Maryland Nanocenter and the NSF MRSEC as a shared experimental facility. We also thank Rich Fiore and Protochips, Inc. for their technical help and for supplying the heating holder and grids. We thank Dr. Wen-An Chiou for his help with the operation of the
References (51)
- et al.
Combust. Flame
(2005) - et al.
Combust. Flame
(2004) Combust. Flame
(2005)Combust. Flame
(2004)- et al.
Combust. Flame
(2006) - et al.
J. Aerosol Sci.
(2002) Acta Mater.
(2003)- et al.
Int. J. Refract Metal Hard Mater.
(2009) J. Alloy. Compd.
(2009)- et al.
Combust. Flame
(2009)
J. Colloid Interface Sci.
Proc. Combust. Inst.
Combust. Flame
Combust. Flame
J. Vac. Sci. Technol., B
Multifunct. Energetic Mater.
J. Phys. Chem. B
Propellants, Explos., Pyrotech.
J. Propul. Power
J. Propul. Power
Adv. Mater.
Nano Lett.
Cited by (134)
Enhanced energy delivery of direct-write fabricated reactive materials with energetic graphene oxide
2024, Combustion and FlameSynthesis of Co(OH)F@Al nanobelt array on various substrates for pyro-MEMS
2023, Chemical Engineering Journal