Altering Reactivity of Aluminum with Selective ... - Wiley Online Library

Full Paper DOI: 10.1002/prep.201200102

Altering Reactivity of Aluminum with Selective Inclusion of Polytetrafluoroethylene through Mechanical Activation Travis R. Sippel,[a] Steven F. Son,[a] and Lori J. Groven*[a]

Abstract: Micrometer-sized aluminum is widely used in energetics; however, performance of propellants, explosives, and pyrotechnics could be significantly improved if its ignition barriers could be disrupted. We report morphological, thermal, and chemical characterization of fuel rich aluminum-polytetrafluoroethylene (70–30 wt-%) reactive particles formed by high and low energy milling. Average particle sizes range from 15–78 mm; however, specific surface areas range from approx. 2–7 m2 g 1 due to milling induced voids and cleaved surfaces. Scanning electron microscopy and energy dispersive spectroscopy reveal uniform distribution of PTFE, providing nanoscale mixing within particles. The combustion enthalpy was found to be 20.2 kJ g 1, though a slight decrease (0.8 kJ g 1) results from extended high energy milling due to a-AlF3 formation. For high energy mechanically activated particles, differential scanning calo-

rimetry in argon shows a strong, exothermic pre-ignition reaction that onsets near 440 8C and a second, more dominant exotherm that onsets around 510 8C. Scans in O2-Ar indicate that, unlike physical mixtures, more complete reaction occurs at higher heating rates and the reaction onset is drastically reduced (approx. 440 8C). Simple flame tests reveal that these altered Al-polytetrafluoroethylene particles light readily unlike micrometer-sized aluminum. Safety testing also shows these particles have high electrostatic discharge (89.9–108 mJ), impact (> 213 cm), and friction (> 360 N) ignition thresholds. These particles may be useful for reactive liners, thermobaric explosives, and pyrolants. In particular, the altered reactivity, large particle size and relatively low specific surface area of these fuel rich particles make them an interesting replacement for aluminum in solid propellants.

Keywords: Aluminum · Polytetrafluoroethylene · Mechanical activation · Energetic composites

1 Introduction Aluminum has become one of the most frequently used metallic fuels in composite propellants, pyrotechnics and explosives, yet its efficient use in these energetics remains challenging for several reasons. One such example is the use of micrometer-sized aluminum in propellants, where aluminum’s high ignition temperature and related particle agglomeration results in lower combustion efficiency and increased two-phase flow losses [1]. To overcome these drawbacks, micrometer-sized aluminum has been replaced with nano-sized aluminum (nAl) in experimental propellants and has resulted in improved performance (e.g., shorter particle burning time [2], reduced agglomerate size [3], decreased ignition delay [4], reduced condensed product size [5], and anticipated increases in propellant heat feedback [6]). However, nAl’s utility is reduced by a higher oxide content and very high surface area (10–50 m2 g 1) that leads to propellant processing issues [2]. It is conceivable that some of the drawbacks of aluminum combustion could be overcome by aluminum particles that are micrometer-sized in scale but contain an additional, intimately mixed component incorporated throughout, resulting in an activated fuel particle. Fluorocarbons as oxidizers for aluminum are of particular interest for inclusion and have been proposed in a variety 286

of applications including reactive liners/fragments, heterogeneous explosives, and submerged flares [7–9]. The success of metal-fluorocarbon reactives can predominantly be attributed to a very high (volumetric and gravimetric) heat release resulting from fluorination instead of oxidation [9, 10]. These benefits have been realized in reactive compositions, where higher performance is seen from use of fluorine-based rather than oxygen-based oxidizers [11]. For applications, where high gas production is desired (such as solid propellants), the approx. 1000 8C lower boiling/sublimation point of most metal fluorides compared to their respective oxides [10] can decrease formation of condensed phase product. Reaction of Al with polytetrafluoroethylene (PTFE) as the oxidizer is of particular interest due to PTFE’s high fluorine content (67 mol-%) and the composition’s [a] T. R. Sippel, S. F. Son, L. J. Groven School of Mechanical Engineering Purdue University 500 Allison Road West Lafayette, IN 47906, USA *e-mail: [email protected] Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/prep.201200102 or from the author.

2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Propellants Explos. Pyrotech. 2013, 38, 286 – 295

Altering Reactivity of Aluminum with Selective Inclusion of Polytetrafluoroethylene

high enthalpy of reaction (9 kJ g 1) [9]. However, one particular drawback of some metal-PTFE reactives (as well as other heterogeneous reactives) is the large diffusion distances present in micrometer-sized mixtures. The issue of diffusion limited combustion has been addressed by several researchers either by significant reduction of reactant particle size through use of nanoparticle reactants (e.g., nAl-nPTFE) [12, 13] or mechanical activation (MA) [7]. The reduced diffusion distance resulting from the use of nanoscale particles dramatically decreases the thermal stimulus required to achieve ignition. Specifically, Osborne and Pantoya [12] show heating of nAl-nPTFE (70– 30 wt-%) mixtures results in an exothermic pre-ignition reaction (PIR) at approx. 430 8C, which is approx. 170 8C below the primary ignition temperature of micrometer scale AlPTFE mixtures. In addition, the significantly higher heat release seen from nAl-nPTFE was attributed to more complete combustion. The use of MA has been successfully applied to many heterogeneous energetics, as it provides a top-down approach to decreasing diffusion distances and altering ignition and reaction behavior [7, 14–17]. With MA (often referred to as arrested reactive milling (ARM) [14]), the milling process is interrupted prior to reaching a critical milling energy dose sufficient to induce self-sustained reaction. The milling yields increased reactant interfacial contact and decreased diffusion distances that can exceed those possible with nanoscale physical mixtures, which can lead to reaction at lower temperatures [14]. For example, Dolgoborodov et al. [7] utilized a custom made vibratory mill to produce mechanically activated Al-PTFE mixtures (15–45 wt-% Al) [7, 18, 19] and in shock propagation experiments reported an increase in reaction propagation velocities of up to 1300 m s 1. However, the effect of varying MA treatment on the produced mixtures (i.e., morphology, specific surface area, and heat of combustion) was not reported. Mechanical activation of equiatomic Al-Ni mixtures by White and co-workers [17] shows milling can reduce exotherm onset by as much as 300 8C. Further, the use of these Al-Ni composite particles in a solid propellant by Reese et al. [20] resulted in reduced metal agglomerate size and decreased particle ignition delay. Inclusion, by MA, of low levels (10 wt-%) of a secondary metal such as Fe, Zn, or Ni in aluminum has also been shown to reduce the ignition temperature and alter the low temperature oxidation process of aluminum [21]. However, for composite propellants the addition of secondary metals may not be advantageous and generally lower specific impulse (Isp) is predicted. This is due to the predominantly condensed phase products formed from intermetallic reactions. The objective of this work is to exploit intimate reactant mixing afforded by low and high energy mechanical activation to produce micrometer scale energetic composite particles with decreased reactant diffusion distances that result in altered ignition and reaction characteristics. The effect of milling time and energy on the resulting particle morphology, phase and energy content, and safety characteristics Propellants Explos. Pyrotech. 2013, 38, 286 – 295

(electrostatic discharge, impact, and friction) is presented. Additionally, we present thermal analysis that details the role of milling on the reaction characteristics in both inert and oxidizing environments. The thermal and morphological traits of the as formed reactive composite particles are also compared to similar nanoparticle mixtures examined elsewhere.

2 Experimental Section For this work fuel-rich mixtures of 70 wt-% Al (35 mm, Valimet H30) and 30 wt-% PTFE (35 mm, Sigma-Aldrich 468096) were selected to (i) allow direct comparison to previous nAl-nPTFE results [12] and (ii) to improve overall safety during inert gas milling and handling of sealed milling containers. Low energy milled composite particles were produced in approx. 3 g batches inside argon (99.997 %) filled, 125 mL HDPE bottles (VWR 414004-156) with a US Stoneware (CV-90116) roller mill rotating at 290 RPM. A charge ratio of 70 was used with 75 wt-% 0.95 cm (McMaster-Carr 9529K19) and 25 wt-% 0.476 cm (McMaster-Carr 9529K13) 440C steel media. For comparative purposes, physical mixtures of Novacentrix 50 nm nAl and Dupont Zonyl (MP1110) nanoscale (200 nm) PTFE (nPTFE) were mixed using the stoichiometry and procedure of reference [12]. High energy Al-PTFE MA particles were produced in approx. 1 g batches inside 30 mL HDPE containers (Cole Parmer EW-62201–01) using a charge ratio of 24 (73 wt-% 0.95 cm, 27 wt-% 0.476 cm media). Milling containers were filled with argon (99.997 %) prior to milling on a SPEX (8000 M) high energy mill using a duty cycle of 1 min ON, 4 min OFF. During milling, the container was cooled using a fan. All milled materials were handled in an argon-filled glove box and were passivated prior to use. This was done by adding enough hexane to fully cover the particles and slowly evaporating the hexane in air. The milling duration (degree of milling treatment) was selected based on the critical milling time required to initiate reaction. The temperature of the milling container was monitored during the milling operation by affixing a K-type precision thermocouple (Omega 5SC-TT-K-36-36) to the exterior of the milling container and recording temperature (Omega OM-EL-USBTC-LCD). The maximum temperature increase was at the end of the 1 min ON cycle (approx. 10 8C rise), but after the 4 min OFF cycle, the measured temperature returned to ambient conditions. There was no deviation from this until the critical milling time was reached (ignition occurred). A Bruker D8-Focus powder X-ray diffractometer (Cu-Ka) was used to analyze composite particles using a scan rate of 2 deg min 1. Scanning electron microscopy (SEM) and energy dispersive spectroscopy (EDS) were conducted using a FEI Quanta 3D-FEG. Particles were also encased in epoxy and sectioned with a Reichert Ultracut E ultramicrotome for imaging of the particle interior. A Micromeritics Tristar 3000 surface area analyzer was used to measure spe-


www.pep.wiley-vch.de

287

Full Paper

T. R. Sippel, S. F. Son, L. J. Groven

cific surface area. The samples (approx. 80 mg) were degassed at 50 8C in ultra-high purity nitrogen for 18 h prior to analysis. Average particle size was assessed using a Malvern Mastersizer 2000 with Hydro 2000 mP dispersant unit with hexane as the dispersing medium. Thermal behavior of 3–10 mg samples was determined with a TA Instruments Q600 DSC-TGA over a temperature range of 100 to 800 8C with heating rates ranging from 5 to 50 K min 1 and 100 mL min 1 flow of either ultra-high purity argon (99.999 %) or a mixture of 20 vol-% O2-Ar. Composite enthalpies of combustion were determined using a Parr 1281 oxygen calorimeter with O2 pressure of 3.10 MPa (450 psi) and a 350 mL chlorine-resistant pressure vessel (Parr 1136CL). Prior to ignition, powders were pressed into 50 mg pellets of 3 mm diameter and approx. 50 % maximum density. Pellets were burned in a custom-made alumina-silicate crucible. For each material, four separate tests were conducted and averaged. The computed “maximum” heat of combustion was determined for compositions in 99 wt-% air using the Cheetah 6.0 [22] equilibrium code. Electrostatic discharge (ESD), impact, and friction sensitivity tests were conducted on 52 h low energy and 60 min high energy MA composite powders. For all sensitivity tests, the Neyer Software [23] was used to determine ignition probability as a function of stimulus strength. Electrostatic discharge testing was conducted on approx. 8 mg powder samples using a custom made apparatus described elsewhere [24]. The ESD machine was operated in oscillatory mode with a 0.1 mF capacitance and variable discharge voltage ranging from 100 to 10,000 VDC. Measurements were made inside an environmental box held at 33 2 % relative humidity by a saturated salt solution. Twenty tests were conducted with each material in order to determine a 50 % ignition threshold. Impact sensitivity experiments were conducted on 10 mg samples using a 5.0 kg weight dropped from various heights. The detailed procedure and test apparatus used are described elsewhere [24]. The MA composite powder was placed on 180-grit sand paper inside a confinement chamber. The chamber pressure was recorded during the test using a PCB (102M232) dynamic pressure transducer and oscilloscope. Ignition was indicated by one or a combination of pressure signal, audible report, and/or presence of combustion products in the chamber. Friction tests were conducted on 3 mg powder samples using a BAM (Bundesanstalt fr Materialforschung) friction tester. Supporting Information (see footnote on the first page of this article): Scanning electron microscopy image of an Al/PTFE (70/30 wt-%, 20 min MA) composite particle as well as a video showing flame ignition of Al/PTFE MA composite powders.

288


3 Results and Discussion While both high and low energy milling were found to be amenable to producing intimately mixed Al-PTFE (70– 30 wt-%) composite particles with reactivity similar to that of nAl-nPTFE physical mixtures, the necessary time for MA was quite different for the two milling methods. High energy milling times in excess of 60 min were sufficient to initiate reaction during milling, while a low energy critical milling time was not reached even at 52 h. This is due to different milling energy doses, which also leads to different critical milling times (time to induce self-sustained reaction). In general, thermal and morphological properties of milled composite particles were repeatable but sensitive to milling conditions – specifically high energy milled materials were sensitive to cooling time and fan speed, as reduction of milling cycle cooling time from 4 to 1 min decreased the critical milling time to approx. 35 min and a similar effect was observed in milling without fan cooling. With 60 min MA, the resulting Al-PTFE composite particles are pyrophoric and require passivation by gradual exposure to air. The specific surface areas of composite particles (Table 1) range from 2.0 to 6.7 m2 g 1 and show that increased cold welding occurs with longer duration high energy milling, resulting in lower specific surface area. This decrease in specific surface area coincides with the increase in average particle size observed from volume weighted particle size distributions obtained from forward light scattering measurements. These results, shown in Figure 1, indicate particle size distributions of milled particles are lognormal and the average particle size of high energy milled materials increases from 55.8 mm (20 min MA) to 78.4 mm (60 min MA). The particle size distributions of high energy milled materials are broad and positively skewed, while the size distribution of low energy milled (52 h) particles is highly uniform with an average particle size of 15.4 mm. Scanning electron microscopy and the significantly smaller average particle size and comparable specific surface area of low energy milled particles revealed that these particles are flake like in morphology and indicated that the surface of low energy milled particles is smoother and contains fewer surface features. The higher specific surface area of high energy milled particles is expected to be a result of the higher energy milling method, which leads to strain hardening of the aluTable 1. Specific surface areas of Al-PTFE (70–30 wt-%) neat and MA composite particles. Material/Milling time

BET SSA/m2 g

Physical mixture 52 h Low energy 20 min High energy 40 min High energy 60 min High energy

0.048 0.025 3.2 0.1 6.7 0.2 5.6 0.1 2.0 0.1


1

50 % ESD ignition threshold/mJ – 108 – – 89.9



Figure 1. Volume-weighted particle size distributions of neat and milled Al-PTFE (70–30 wt-%) composite particles.

minum matrix and reduced cold welding efficiency at longer milling times [25]. Effects typical of strain hardening were also observed in SEM images of a high energy MA particle (Figure 2a), where incomplete cold welding produces voids, cleaved surfaces, and incompletely consolidated flakes on the particle surface. While individual particles remain in the range of 20–300 mm, at 60 min MA, the decreased aluminum cold welding efficiency results in a highly cleaved surface and visible pockets. SEM images of 52 h low energy (Figure 2c) and 20 min high energy MA composites (see Supporting Information) indicate that cold welding and subsequent strain hardening were less pronounced at lower MA times as well as for low energy MA. In the initial stages of milling, we expect the milling mechanism responsible for forming these composite particles to be typical of ductile-ductile milling [25]. During this process, the more ductile material

(PTFE) deforms and coats the higher yield strength material (aluminum), minimizing exposure of unoxidized metal surfaces and reducing material specific surface area. As particles are cold welded together, alternating lamellar layers of PTFE and aluminum form within particles, resulting in high reactant interfacial area. With continued milling, aluminum strain hardening occurs and the PTFE appears frayed into 10–50 nm diameter PTFE fibers. These fibers are evident in Figure 2b (black arrows), which shows the interior of a 60 min MA particle. The intimate mixing of aluminum and PTFE is apparent from EDS of a high energy milled (60 min) particle, shown in Figure 3. Elemental analysis shows even distribution of fluorine throughout the particle’s aluminum matrix. It is worth noting that at an accelerating voltage 20 kV, localized ignition of high energy MA particles occurred within the microscope. Although MA composite particles ignited inside the microscope, safety characterization indicates that the particles are relatively insensitive to ESD, impact, and friction. The 50 % ESD ignition threshold of 52 h low energy and 60 min high energy MA composite powders (shown in Table 1) are 108 and 89.9 mJ, respectively. For comparison, the 50 % ESD ignition threshold for 80 nm nAl is 5.0 mJ [26]. Ignition thresholds for impact and friction tests were not determined, as ignition was not observed at the highest stimulus level (213 cm and 360 N, respectively). Ten tests were conducted at this stimulus level and all resulted in failed ignition. X-ray diffraction of milled, neat, and physically mixed materials (Figure 4) indicates substantial peak broadening as a result of both crystallite size reduction and milling induced strain. Scherrer analysis of peaks indicates 60 min high energy MA reduces aluminum crystallite size from 59 to 24 nm and PTFE crystallite size from 26 to 9 nm. With extended high energy milling we observe gradual formation of a-AlF3 but no milling-induced Al4C3 formation. This gradual formation of product species has been observed in high energy milling of other reactive mixtures [27, 28] and may be caused by milling-induced reactions that occur locally at milling impact sites. However, the low impact energy of roller milling appears to be insufficient to produce detectable quantities of intermediates, as no product

Figure 2. SEM micrographs of (a) single 60 min high energy MA particle, (b) 60 min high energy MA particle interior structure and inset detail, and (c) single 52 h low energy MA particle. Black arrows indicate PTFE fibers.




289

Full Paper


Figure 3. SEM micrograph of 60 min high energy (SPEX) milled Al-PTFE (70–30 wt-%) particle (left) and EDS elemental map (right) of the particle showing presence of aluminum (Al), fluorine (F), and carbon (C).

Figure 5. Enthalpy of combustion of Al-PTFE low energy (LE) and high energy (HE) MA particles and nanoparticle mixtures. Error bars indicate the standard deviation of four tests.

Figure 4. XRD patterns of Al-PTFE (70–30 wt-%) MA particles and physical mixtures.

species were detected by XRD. Although presence of aAlF3 in high energy MA composites suggests a reduction in the energy content, oxygen calorimetry (Figure 5) of these materials indicates the degree of a-AlF3 formation and subsequent energy reduction is minor. Increasing milling time from 20 to 60 min (high energy MA) decreases composite particle enthalpy of combustion from 19.0 0.5 to 18.2 0.8 kJ g 1. The overall higher heat release of low energy milled materials (20.2 0.8 kJ g 1) further suggests that aAlF3 formation is in part responsible for the slight reduction in heat release resulting from longer duration and higher intensity milling. The formation of some Al2O3 in these fuel rich composite particles is also expected due to initial exposure of the ma290


terial to air after the MA process. However this Al2O3 is not detected by XRD due to its amorphous nature. Its presence, however, results in a decrease in combustion enthalpy from the maximum, computed value of 24.3 kJ g 1 to that of low energy MA composites (20.2 0.8 kJ g 1). Successive air aging of low energy MA composite particles for 100 days further reduces combustion enthalpy to 19.4 0.9 kJ g 1 (Figure 5). Perhaps the most noticeable difference in combustion enthalpies, shown in Figure 5, is between that of MA composites and similar nAl-nPTFE physical mixtures. Due to the lower aluminum oxide content of MA composites, the computed enthalpy of combustion of MA composite particles (0 wt-% Al2O3, ~Hc = 24.3 kJ g 1) is about 30 % higher than the computed enthalpy of combustion of nAlnPTFE physical mixtures (25 wt-% Al2O3, ~Hc = 18.9 kJ g 1) and nearly 95 % higher than the measured nAl-nPTFE combustion enthalpy (12.5 0.2 kJ g 1). The difference between measured nAl-nPTFE enthalpy of combustion and the computed value could be due to a combination of manufacturer batch variation, poor mixing of nAl-nPTFE mixtures during sonication and drying, or settling of mixtures during handling. In addition to MA composites having combustion enthalpies higher than nAl-nPTFE, the MA process alters reactivity




Figure 6. DSC (20 K min 1, argon) heat flow (left) and sample weight history (right) of Al-PTFE (70–30 wt-%) reactive composite compared to results of Osborne and Pantoya [12]. Heat flow signals shifted 15 W g 1 and weight signals shifted 20 % for presentation.

from that of micrometer-sized precursor mixtures, resulting in materials with ignitability and reaction characteristics similar to those of nAl-nPTFE physical mixtures without the drastic energy reduction or high surface areas. Simple flame tests reveal that the MA process alters ignitability, as all Al-PTFE MA composites ignite readily upon application of a butane flame, while ignition of physical mixtures of micrometer-sized precursor powders are only ignitable with continued flame exposure. Digital video of these flame tests is available from the publisher’s website. To elucidate composite particle reactivity and gain insight into their ignition characteristics, DSC-TGA experiments were conducted to compare composite reaction with that of unmilled precursor and nAl-nPTFE mixtures. First, we consider just the reaction of Al-PTFE particles by analysis in an argon atmosphere (Figure 6). In heating of micrometer-sized physical mixtures we initially observe melting of PTFE near 327 8C followed by PTFE decomposition (onset approx. 500 8C) and sample weight loss. This first step of the Al-PTFE reaction is decomposition of PTFE into gaseous products and was reported by Turetsky et al. [29] and also predicted by Losada and Chaudhuri [30]. Decomposition of PTFE occurs rapidly until 600 8C, at which point nearly all the PTFE (27 % of sample weight) has decomposed. As PTFE decomposition ceases (615 8C), a weak exotherm occurs and finally at 660 8C melting of unreacted aluminum occurs. In micrometer-sized physical mixtures, Propellants Explos. Pyrotech. 2013, 38, 286 – 295

only a small portion of PTFE reacts with aluminum, as is evident by the weak exotherm and prominent aluminum melt endotherm. This is a result of a lack of reaction interfacial area, as slightly greater exothermicity occurs in reaction of micrometer-sized aluminum with nPTFE observed by Osborne and Pantoya [12] as shown in Figure 6. However, reaction of both of these mixtures is exclusively limited to temperatures of 550–640 8C, where PTFE decomposition occurs. The degree of reaction occurring in these mixtures is also low, as aluminum melt endotherms are prominent. This is not the case for MA composite particles, which undergo a reaction that is more representative of nAlnPTFE (Figure 6) in which Osborne and Pantoya [12] observe the occurrence of a pre-ignition reaction (PIR) at approx. 430 8C followed by a primary exotherm at approx. 540 8C. Considering first the low energy MA composite particles, we initially observe PTFE melting at 327 8C. During low temperature heating of composite particles, we expect interparticle strain to occur, as the coefficient of linear thermal expansion of PTFE is about 10 times higher than that of aluminum. In heating from room temperature to melting temperature (327 8C), PTFE volumetrically expands by 36 % [31], causing particles to strain, exposing unoxidized aluminum surfaces. Following PTFE melting, exothermic PIR reaction onsets at 440 8C, which is far below the reaction temperature of micrometer-sized physical mixtures. This PIR reaction occurs in the condensed phase without significant



291

Full Paper


Figure 7. DSC heat flow (left) and sample weight history (right) of Al-PTFE (70–30 wt-%, 20 vol-% O2-Ar) composite particles and 35 mm neat aluminum and 35 mm neat PTFE. Heat flow signals shifted 20 W g 1 and weight signals shifted 20 % for presentation.

weight loss and is a result of exothermic fluorination of alumina [13, 32]. Exothermic fluorination is immediately followed by rapid weight loss caused by PTFE decomposition. During this process, PTFE product gases generated throughout composite particles may raise the pressure inside the particles, further increasing particle stress until aluminum and PTFE surfaces debond, allowing PTFE decomposition gases to react at aluminum surfaces and to escape. Due to the high milling-induced interfacial surface area within particles, reaction can occur much faster than in micrometer-sized mixtures and leads to more efficient use of PTFE decomposition products. Additionally, the reaction rate is increased by the higher species diffusivity caused by milling [25]. Subsequently, a second (primary) exotherm onsets near 520 8C. This exotherm is believed to be initiated by two simultaneous, exothermic phase transformations, in which amorphous Al2O3 is converted to gAl2O3 [33, 34] and b-AlF3 to a-AlF3 [32]. During the onset of these two phase transformations, heat release causes decomposition of remaining PTFE and successive reaction with aluminum. Aluminum fluorination may be further facilitated by the exposure of aluminum surfaces due to breakup of the Al2O3 passivation layer caused by densification of Al2O3 in transition from the amorphous to g-phase [33]. A similar two-step exothermic behavior is observed in the heating of high energy MA composites. An exothermic PIR reaction onsets at approx. 440 8C accompanied by a 5 % 292


sample weight loss resulting from PTFE decomposition. The PIR reaction occurs and is followed by a main exotherm that onsets at approx. 510 8C. However, the onset temperatures of the PIR and main exotherm vary slightly from those observed from low energy MA composites due to the varying degree of intermixing caused by the different milling conditions. Additionally, the magnitude of the high energy MA composite PIR is substantially greater than that of low energy MA composites. Following the PIR and main exotherm, a weak aluminum melting endotherm occurs at 660 8C and finally, an additional, weak, “late second exotherm” (approx. 740 8C) that is believed to be aluminum oxide phase transformations from g-Al2O3 to d-Al2O3 and/or q-Al2O3 [32, 34]. While DSC experiments in argon allow us to assess MA effects on Al-PTFE interaction, experiments in the presence of an additional oxidizer species are more representative of the environment (e.g., composite propellants, enhanced blast, etc.), in which these fuel rich (70 wt-% Al) particles will be used. Therefore, additional DSC-TGA experiments were conducted at various heating rates in the presence of 20 vol-% O2-Ar. In DSC heating of physical, micrometersized mixtures at 20 K min 1 (Figure 7), an exotherm and corresponding rapid sample weight loss occurs around 530–580 8C, which is caused by PTFE decomposition and reaction with oxygen. This is confirmed by heating neat PTFE in O2-Ar and is consistent with the reaction mechanism proposed by Losada and Chaudhuri [30] that showed Al-PTFE




reaction pathways beginning with O2-PTFE decomposition species are more favorable (lower activation energy, higher exothermicity) and faster than anaerobic pathways. Consequently, in the case of DSC heating of physical Al-PTFE mixtures, we expect nearly all observed heat release is attributed to PTFE decomposition products reacting with oxygen and any Al-PTFE interaction is obscured. This leads to a strong aluminum melting endotherm at 660 8C, which is approximately the same magnitude as the melt endotherm caused from the heating of neat aluminum. As shown in Figure 7, heating behavior of low energy MA composites is similar to physical mixtures but is more exothermic. However, in low energy MA composites (20 K min 1), the exotherm temperature is decreased to 500–580 8C due to the intimate mixing afforded by low energy MA. In contrast to low energy MA particles, high energy MA (60 min) particles exhibit far different behavior when heated in O2-Ar (Figure 7). Upon heating (20 K min 1), we observe a broad, low temperature exotherm (which onsets at 225 8C). This heat release is likely due to some HDPE contamination from the milling container, as this behavior is not observed when milling is conducted in polypropylene containers. A second exotherm onsets at approx. 460 8C that corresponds to the previously described PIR. This exotherm is accompanied by an 8 % sample weight loss that is likely due to both PTFE decomposition and exothermic reaction of decomposition products with aluminum and oxygen. A third exotherm accompanied by sample weight gain broadly onsets near 550 8C and is initiated by the two exothermic Al2O3 and AlF3 phase transitions observed in argon DSC and previously discussed. This heat release, which is a result of PTFE decomposition products and oxygen reacting with aluminum, greatly accelerates during the melting of aluminum and peaks at 660 8C. At this point, near complete reaction of aluminum is indicated by the lack of an aluminum melt endotherm. At a 50 K min 1 heating rate, a broad, low temperature exotherm is also observed at 225 8C. At this heating rate, the first major exotherm onset corresponds to the previously described PIR (approx. 440 8C). This reaction results in near complete aluminum oxidation (and greater heat release) as is evident from the corresponding 10 % weight gain and a weak aluminum melting endotherm observed at 660 8C. Aluminum melting is followed by a late second exotherm and further weight gain (oxidation) of 7 %. The maximum heat flow from high energy MA composite particles (approx. 100 W g 1) is substantially higher than physical mixtures or low energy MA particles (approx. 20 W g 1) at 50 K min 1. In addition to higher exothermicity, the absence of aluminum melting endotherm in the heating of high energy MA composite particles at 20 K min 1 indicates a greater extent of aluminum reaction. Furthermore, comparison of the heating of high energy MA composites to that of 35 mm neat aluminum particles shows the drastically modified behavior of aluminum combustion caused by MA of these fuel rich composite particles. Propellants Explos. Pyrotech. 2013, 38, 286 – 295

4 Conclusions This work shows that MA of fuel rich Al-PTFE mixtures can result in micrometer-sized Al-PTFE composite particles with increased reactivity. We have shown that use of MA results in nanoscale mixing of reactants with reaction behavior similar to that of nAl-nPTFE. Notably, high or low energy MA results in significant reduction of primary exotherm onset from 600 8C to 440 8C in anaerobic heating and from 540 8C to 440 8C in presence of O2. For composite particles formed with high energy MA, differential scanning calorimetry in O2-Ar indicates that, unlike physical mixtures or those particles formed under low energy MA, more complete reaction occurs. At higher heating rates the reaction onset is also drastically reduced (approx. 440 8C). Furthermore, results suggest that at aerobic heating rates greater than 50 K min 1, near complete heat release occurs by approx. 600 8C instead of at higher temperatures. While MA drastically alters the reactivity of these particles, they are relatively insensitive to ESD, impact, and friction initiation. In addition to having significantly altered reaction behavior, the enthalpy of combustion of MA particles was found to be as high as 20.2 kJ g 1, which is approximately 60 % higher than the measured combustion enthalpy of nAlnPTFE mixtures. Additionally, the large (15 to 78 mm) average particle size and moderate specific surface areas (2 to 6.7 m2 g 1) of composite particles suggests they will be far more useful than nanoparticles in high solids loaded energetics and may age more favorably than nanoparticle mixtures. We expect that further reduction of particle specific surface area and improvement of aging characteristics may be achieved by adding a small amount of binder (e.g., Viton A) during the milling process or through crash deposition after MA particle formation. A lower fraction of PTFE may also prove to be advantageous for some applications. These micrometer-sized activated fuel particles with altered ignition and reaction characteristics are a promising alternative to nanoparticle solid propellant additives such as nAl. With these particles, we expect similar propellant performance increases can be achieved with less detriment to propellant mechanical and rheological properties. When used as a replacement in solid propellants, these particles may ignite far below the ignition temperature of micrometer-sized aluminum (> 2000 8C) and we expect they may decrease ignition delay, agglomerate size, and reduce condensed phase losses as well as lead to increased heat release and higher burning rates. Use of these fuel rich AlPTFE composite particles in structural energetics (e.g., reactive liners), flares, incendiaries, and other energetics could also likely lead to performance far exceeding that of energetics made from physical mixtures of micrometer- or nanometer-sized particles. Efforts are now focused on the use of other fluorocarbon oxidizers and study of the ignition and combustion of these activated fuel particles at high heating rates. Additional work to incorporate these materials into solid and



293

Full Paper


hybrid propellants and structural reactives is presently underway. The effects of such addition on solid propellant combustion are also being studied.

Symbols and Abbreviations ~Hc

ARM BAM DSC EDS ESD Isp HDPE MA nAl nPTFE PIR PTFE SEM TGA XRD

Enthalpy of combustion Arrested reactive milling Bundesanstalt fr Materialforschung Differential scanning calorimetry Energy dispersive spectroscopy Electrostatic discharge Specific impulse High density polyethylene Mechanical activation Nanoaluminum Nanoscale polytetrafluoroethylene Pre-ignition reaction Polytetrafluoroethylene Scanning electron microscopy Thermogravimetric analysis X-ray diffraction

Acknowledgments The authors would like to acknowledge the financial support of the Air Force Office of Scientific Research under the supervision of Dr. Mitat Birkan (#FA9550-09-1-0073). We would also like to thank Debby Sherman for conducting the SEM/EDS and Mario Rubio for his help conducting safety testing.

References [1] R. L. Geisler, A Global View of the Use of Aluminum Fuel in Solid Rocket Motors, 38th AIAA/ASME/ASE/ASEE Joint Propulsion Conference and Exhibit, Indianapolis, IN, USA, July 7–10, 2002, AIAA 2002–3748. [2] R. A. Yetter, G. A. Risha, S. F. Son, Metal Particle Combustion and Nanotechnology, Proc. Combust. Institute 2009, 32, 1819 – 1838. [3] K. Jayaraman, S. R. Chakravarthy, R. Sarathi, Quench Collection of Nano-Aluminium Agglomerates from Combustion of Sandwiches and Propellants, Proc. Combust. Institute 2011, 33, 1941 – 1947. [4] V. A. Arkhipov, A. G. Korotkikh, The Influence of Aluminum Powder Dispersity on Composite Solid Propellants Ignitability by Laser Radiation, Combust. Flame 2012, 159, 409 – 415. [5] A. Dokhan, E. W. Price, R. K. Sigman, J. M. Seitzman, The Effects of Al Particle Size on the Burning Rate and Residual Oxide in Aluminized Propellants, 37th AIAA/ASME/ASE/ASEE Joint Propulsion Conference and Exhibit, Salt Lake City, UT, USA, July 8–11, 2001, AIAA Paper 2001–3581. [6] A. Dokhan, E. W. Price, J. M. Seitzman, R. K. Sigman, The Effects of Bimodal Aluminum with Ultrafine Aluminum on the Burning Rates of Solid Propellants, Proc. Combust. Institute 2007, 29, 2939 – 2946.

294


[7] A. Y. Dolgoborodov, M. N. Makhov, I. V. Kolbanev, A. N. Streletskii, V. E. Fortov, Detonation in an Aluminum-Teflon Mixture, JETP Lett. 2005, 81, 311 – 314. [8] E. C. Koch, Review on Pyrotechnic Aerial Infrared Decoy Flares, Propellants Explos. Pyrotech. 2001, 26, 3 – 11. [9] E. C. Koch, Metal-Halocarbon Pyrolant Combustion in: Handbook of Combustion: New Technologies (Eds.: M. Lackner; F. Winter; A. K. Agarwal), Wiley-VCH: Weinheim, 2010, pp. 355 – 365. [10] C. M. Beighley, W. R. Fish, R. E. Anderson, Advanced Fuels and Oxidizers in: Propellants Manufacture, Hazards, and Testing (Eds.: C. Boyars; K. Klager), American Chemical Society: Washington DC, 1969, pp. 319 – 320. [11] K. W. Watson, M. L. Pantoya, V. I. Levitas, Fast Reactions with Nano- and Micrometer Aluminum: A Study on Oxidation vs. Fluorination, Combust. Flame 2008, 155, 619–634. [12] D. T. Osborne, M. L. Pantoya, Effect of Al Particle Size on the Thermal Degradation of Al/Teflon Mixtures, Combust. Sci. Technol. 2007, 179, 1467–1480. [13] M. L. Pantoya, S. W. Dean, The Influence of Alumina Passivation on Nano-Al/Teflon Reactions, Thermochim. Acta 2009, 493, 109 – 110. [14] E. L. Dreizin, Metal-Based Reactive Nanomaterials, Prog. Energy Combust. Sci. 2009, 35, 141 – 167. [15] M. Schoenitz, T. S. Ward, E. L. Dreizin, Preparation of Energetic Metastable Nanocomposite Materials by Arrested Reactive Milling, Mater. Res. Soc. Proc. 2003, 800, 1 – 6. [16] M. Schoenitz, T. S. Ward, E. L. Dreizin, Fully Dense Nano-Composite Energetic Powders Prepared by Arrested Reactive Milling, Proc. Combust. Institute 2005, 30, 2071 – 2078. [17] J. D. E. White, R. V. Reeves, S. F. Son, A. S. Mukasyan, Thermal Explosion in Al-Ni System: Influence of Mechanical Activation, J. Phys. Chem. A 2009, 113, 13541–13547. [18] A. N. Streletskii, A. Y. Dolgoborodov, I. V. Kolbanev, M. N. Makhov, S. F. Lomaeva, A. B. Borunova, V. E. Fortov, Structure of Mechanically Activated High-Energy Al Plus Polytetrafluoroethylene Nanocomposites, Colloid J. 2009, 71, 852 – 860. [19] A. Y. Dolgoborodov, A. N. Streletskii, M. N. Makhov, I. V. Kolbanev, V. E. Fortov, Explosive Compositions Based on the Mechanoactivated Metal-Oxidizer Mixtures, Russ. J. Phys. Chem. B 2007, 1, 606 – 611. [20] D. A. Reese, L. J. Groven, S. F. Son, A. S. Mukasyan, Intermetallic Compounds as Fuels for Composite Rocket Propellants, 47th AIAA/ASME/ASE/ASEE Joint Propulsion Conference and Exhibit, Fort San Diego, CA, USA, July 31–August 3, 2011, AIAA 2011– 5865. [21] Y. Aly, M. Schoenitz, E. L. Dreizin, Aluminum-Metal Reactive Composites, Combust. Sci. Technol. 2011, 183, 1107 – 1132. [22] L. E. Fried, W. M. Howard, P. C. Souers, P. A. Vitello, Cheetah 6.0 User Manual, Report LLNL-SM-416166, Lawrence Livermore National Laboratory, Livermore, CA, USA, 2010. [23] B. T. Neyer, Sentest 1.0 Software Manual, Neyer Software, Cincinnati, OH, USA, 2000. [24] T. R. Sippel, S. F. Son, G. A. Risha, R. A. Yetter, Combustion and Characterization of Nanoscale Aluminum and Ice Propellants, 44th AIAA/ASME/ASE/ASEE Joint Propulsion Conference and Exhibit, Hartford, CT, USA, July 20–23, 2008, AIAA 2008–5040. [25] C. Suryanarayana, Mechanical Alloying and Milling, Prog. Mater. Sci. 2001, 46, 1 – 184. [26] T. R. Sippel, T. L. Pourpoint, S. F. Son, Combustion of Nanoaluminum and Water Propellants: Effect of Equivalence Ratio and Safety/Aging Characterization, Propellants Explos. Pyrotech. 2012, accepted.



Altering Reactivity of Aluminum with Selective Inclusion of Polytetrafluoroethylene [27] L. Takacs, Metal-Metal Oxide Systems for Nanocomposite Formation by Reaction Milling, NanoStructured Mater. 1993, 2, 241 – 249. [28] L. Takacs, Self-Sustanining Reactions Induced by Ball Milling, Prog. Mater. Sci. 2002, 47, 355 – 414. [29] A. L. Turetsky, F. Block, G. Young, High Temperature Thermal Analysis of a Pyronol Composition, Thermochim. Acta 1992, 212, 197 – 207. [30] M. Losada, S. Chaudhuri, Theoretical Study of Elementary Steps in the Reactions between Aluminum and Teflon Fragments under Combustive Environments, J. Phys. Chem. A 2009, 113, 5933 – 5941. [31] P. Zoller, The Specific Volume of Poly(Tetrafluoroethylene) as a Function of Temperature (30 8C-372 8C) and Pressure (0– 2000 kg/cm2), J. Appl. Polym. Sci. 1978, 22, 633–641.


[32] Z. Sarbak, Effect of Fluoride and Sodium Ions on Structural and Thermal Properties of g-Al2O3, Cryst. Res. Technol. 1997, 32, 491–497. [33] M. A. Trunov, M. Schoenitz, X. Zhu, E. L. Dreizin, Effect of Polymorphic Phase Transformations in Al2O3 Film on Oxidation Kinetics of Aluminum Powders, Combust. Flame 2005, 140, 310– 318. [34] O. Schulz, N. Eisenreich, H. Feietzek, B. Eickershoff, M. Schneider, E. Kondratenko, Structural Changes During the Oxidation of Micrometer-Sized Al Particles up to 1523 K in Air, Z. Kristallogr. Proc. 2011, 1, 367–372.


Received: June 25, 2012 Revised: August 31, 2012


295