COMBUSTION BEHAVIORS OF BIMODAL ALUMINUM
SIZE DISTRIBUTIONS IN THERMITES
by
KEVIN MOORE, B.S.M.E.
A THESIS
IN
MECHANICAL ENGINEERING
Submitted to the Graduate Faculty of Texas Tech University in
Partial Fulfillment of the Requirements for
the Degree of
MASTER OF SCIENCE
IN
MECHANICAL ENGINEERING
Approved
Michelle Pantoya Chairperson of the Committee
Brandon Weeks
Louisa Hope-Weeks
Accepted
John Borrelli Dean of the Graduate School
May, 2005
ii
ACKNOWLEDGMENTS
First of all, I would like to thank my masters supervisor Dr. Michelle Pantoya for
all of the hard work she has dedicated to me reaching this point in my academic career. I
am so very grateful that she took a chance on me entering her research program, which
has been integral in my development as a student. Her positive attitude and optimistic
approach to life were crucial to my success.
I would also like to thank my supervisor and mentor at Los Alamos National
Laboratory, Dr. Steve Son. I am grateful for the opportunity to work with him and gain
so much experience from his guidance. He was always willing to take time out to talk
with me about whatever is on my mind, for which I am very thankful.
I would like to gratefully acknowledge the support of the Army Research Office
(Contract Number DAAD19-02-1-0214) and our program manager, Dr. David Mann.
I would also like to acknowledge support by the Los Alamos National Laboratory
through the Advanced Energetics Initiative, the Defense Threat Reduction Agency
(DTRA), and the Seaborg Institute.
I would like to thank the Texas Tech University Honors College for their support
through the Undergraduate Research Fellowship Program.
I would like to thank my wife, Lindsay, for the love and support she has given me
through this trying period. I would like to also acknowledge my family for their loving
support and always motivating me to put my best foot forward.
I would also like to thank several co-workers that I have had during my work
towards this point. Bryan Bockmon was an excellent mentor, during the Texas Tech
iii
Mechanical Engineering Mentoring Program in the spring of 2002, and first introduced
me to the world of energetic materials. I learned so much from Jim Busse and Eric
Sanders, co-workers and lab mentors at Los Alamos National Laboratory, and were an
immense help in performing my summer experiments. Also, I would like to thank John
Granier for all of his help developing experiments here at Texas Tech.
iv
TABLE OF CONTENTS
ACKNOWLEDGMENTS .................................................................................................. ii
ABSTRACT....................................................................................................................... vi
LIST OF TABLES............................................................................................................ vii
LIST OF FIGURES ......................................................................................................... viii
NOMENCLATURE ........................................................................................................... x
CHAPTER
I. INTRODUCTION............................................................................................. 1
1.1 Thermites ................................................................................................ 1 1.2 Nano-composite Thermites..................................................................... 2 1.3 Aluminum Combustion........................................................................... 4 II. EXPERIMENTAL ............................................................................................ 7
2.1 Sample Preparation ................................................................................ 7 2.2 Laser Ignition of Pellets ....................................................................... 10 2.3 Laser Ignition of Loose Powders ......................................................... 11 2.4 Combustion Velocities of Loose Powders........................................... 12 2.5 Pressure Cell Tests............................................................................... 13 III. THEORY ........................................................................................................ 14
3.1 Flame Structure..................................................................................... 14 3.2 Diffusion vs. Chemical Kinetics .......................................................... 15 IV. RESULTS....................................................................................................... 18
4.1 Ignition Sensitivity............................................................................... 18
v
4.2 Combustion Velocity ........................................................................... 21 V. DISCUSSION ................................................................................................. 24
5.1 Ignition Sensitivity............................................................................... 24 5.2 Combustion Velocity ........................................................................... 26 VI. CONCLUSIONS ............................................................................................ 36
VII. IMPLICATIONS AND FUTURE RESEARCH............................................ 38
7.1 Nano-scale Aluminum Properties ........................................................ 38 7.2 Mixing Techniques .............................................................................. 39 7.3 Preliminary Molybdenum Trioxide Study........................................... 40 REFERENCES ................................................................................................................. 47
vi
ABSTRACT
In recent years many studies that incorporated nano-scale or ultrafine aluminum
(Al) as part of an energetic formulation demonstrated significant performance
enhancement. Decreasing the fuel particle size from the micron to nanometer range alters
the material’s chemical and thermal-physical properties. The result is increased particle
reactivity that translates to an increase in the combustion velocity and ignition sensitivity.
Little is known, however, about the critical level of nano-sized fuel particles needed to
enhance the performance of the energetic composite. Ignition sensitivity and combustion
velocity experiments were performed using a thermite composite of Al and molybdenum
trioxide (MoO3) at the theoretical maximum density (TMD) of a loose power (5% TMD)
and a compressed pellet (50% TMD). A bimodal Al particle size distribution was
prepared using 4 or 20 µm Al fuel particles that were replaced in 10% increments by 80
nm Al particles until the fuel was 100% 80 nm Al. These bimodal distributions allow the
unique characteristics of nano-scale materials and their interactions with micron scale Al
particles to be better understood.
vii
LIST OF TABLES
1. Reactant Particle Description................................................................................................7 2. Calculated thermal and physical parameters for all Al-MoO3 composites ........................28
viii
LIST OF FIGURES
1. A SEM micrograph of an 80 nm and 4 µm Al mixture ..................................................8 2. Schematic diagram illustrating aluminum distribubimodal mixture ratios.....................9 3. Schematic diagram of the unconfined loose powder ignition sensitivity test
apparatus. .....................................................................................................................12 4. Schematic diagram of the combustion velocity test apparatus for loose powders. .....12 5. Typical thermite flame structure diagram.....................................................................14 6. Still frame images of ignition and flame propagation of an Al-MoO3 pellet with
an Al distribution of 60 wt % 80 nm and 40 wt % 4 µm.. ...........................................18 7. Ignition time for pellets as a function of percent 80 nm Al content mixed with
MoO3 and 4 µm Al at 50 W laser power ....................................................................19 8. Ignition time for pellets as a function of percent 80 nm Al content mixed with
MoO3 and 20 µm Al at 50 W laser power ..................................................................20 9. Ignition time for loose powders as a function of percent 80 nm Al content mixed
with MoO3 and 4 or 20 µm Al at 50 or 100 W laser power.........................................20 10. Combustion velocity as a function of weight percent 80nm Al content for pellets
with 4 µm Al and MoO3 .............................................................................................22 11. Combustion velocity as a function of weight percent 80nm Al content for pellets
with 20 µm Al and MoO3 ...........................................................................................22 12. Combustion velocity as a function of weight percent 80nm Al content for loose
powder with 4 or 20 µm Al and MoO3.......................................................................23 13. Pressure output as a function of percent nano-Al content for loose powder
Al-MoO3 mixtures with 4µm and 20µm Al. ..............................................................29 14. Combustion of 100% 80nm Al and MoO3 loose powder mixture..............................31 15. Combustion of loose powder Al and MoO3 mixture with 10% 80nm Al and
90% 20µm Al .............................................................................................................32 16. Combustion of 100% 4µm Al and MoO3 loose powder mixture................................33
ix
17. SEM images of products of 50% 80 nm Al and 50% 20 µm Al pellet mixtures
with MoO3....................................................................................................................34
NOMENCLATURE
Variable
A Arrhenius pre-exponential factor
B Mass transfer number
C Species concentration
D Mass diffusivity
Ea Activation energy
k Surface reaction rate constant
"m& Mass flux
r Radial dimension
R Gas constant
R Spherical fuel particle outer radius
T Temperature
Y Mass fraction
Greek
ρ Pure fuel density
ω Reaction rate
x
1
CHAPTER I
INTRODUCTION
1.1 Thermites
A thermite reaction is often described as a reaction between a metal and a metallic
or non-metallic oxide, which results in a more stable metallic oxide and the
corresponding metal or non-metal of the reactant oxide [1]. An important product of
these reactions is a high heat of reaction, commonly referred to as energy density [1].
Once ignited, the thermite reaction is self-sustaining due to this highly exothermic heat
release. Thermite reactions are typically thought to be conductively driven reactions.
Most solid energetic materials, including explosives like trinitrotoluene (TNT)
and nitroglycerine, are composed of many molecules of the same engineered material.
These materials are referred to as monomolecular energetic materials. Within each of
these molecules are fuel, typically carbon, and oxidizing molecules which allow the
reaction to occur without the presence of additional oxygen from the environment.
Monomolecular energetics have kinetically controlled reactions which can produce very
high deflagrations or even detonations. Unlike monomolecular energetics, thermites are
mixtures of fuel and oxidizing particles, typically ranging from 1 to 100 microns in size.
Due to the inherent separation of the fuel and oxidizing particles, the reaction is diffusion
controlled, which usually translates to much slower deflagrations. Although thermites
have relatively high energy densities, the slow burning rate causes the energy release rate
to be much slower than that of monomolecular energetics.
2
This slow energy release rate has limited the utility of thermite materials mostly
to areas outside of explosives. Welding underwater or in hard to reach places is often
solved by the use of thermites; molten metals from the reaction can seep between the
metal pieces and form a bead during the cooling process [1]. Self propagating high
temperature synthesis (SHS) is another important application of thermite-type materials
[3-6]. SHS, which often involves the intermetallic reaction between two or more metals,
allows the development of materials with more mechanically favorable microstructures
for specific applications [3]. One-such application is the formation of a ceramic layer on
the inside of metal piping induced by the reaction of a thermite on the inner surface of the
pipe [1].
1.2 Nano-composite Thermites
Particle size reduction of fuel and oxidizing particles within the micron range in
thermite systems has resulted in significantly improved combustion performance.
Shimizu and Saitou [7] have experimentally shown that the reaction rate of the Fe2O3-
V2O5 system increased as the number of contact points between the fuel and oxidizing
particles increased. Brown et al. [8] have seen similar results in the Sb-KMnO4 system.
As the calculated number of contact points in this system doubled due to smaller particle
sizes, the burning rate of the mixtures tripled [8]. Tomasi and Munir [9] have shown that
as the particle size of the Nb2O5-Al2Zr system decreased from 25 to 2 microns, the
reaction burn rate increased by a factor of five. Drennan and Brown [10] have shown
burn rates double as the molybdenum particle size of MoBaO2 and MoSrO2 systems is
reduced from 35 microns to 15 microns. It is apparent that the particle size of the mixture
3
plays a significant role in the combustion performance for all of these systems. The
increase in reactivity for these cases is attributed to a better mixing between fuel and
oxidizing particles. Smaller particle sizes allow the fuel and oxidizing particles to
effectively be closer together, creating smaller diffusion distances that must be overcome
during the reaction.
Recent developments in particle formation technology have created the cost
efficient production of nano-scale particles [11]. One potential for nano-particles to
increase combustion performance is due to diffusion distances decreasing for particles of
this size. In addition to new particle formation technologies, new characterization
technologies, like small angle neutron scattering, have been developed to determine
particle size distributions more accurately for particles in the micron range [12-13].
The reduction of aluminum particle size to the nano-scale in the Al-MoO3 system
has shown promising results. An initial study by Son et al. [14] shows increased burning
rates in nano-sized aluminum thermite mixtures; an important finding of their
experiments shows that radiation is of little importance and that convective heat transfer
during the reaction process is a more prominent factor. Bockmon et al. [15] have shown
that burning rates of Al-MoO3 nano-powder mixtures can reach confined burning rates of
up to 1 km/s. These speeds, which are attributed to a prominent convective burning, are
much faster than thermites are traditionally known to burn. Traditional thermites burn on
the order of 1 m/s [1].
Direct comparison of nano and micron aluminum particle sizes in Al-MoO3
mixtures by Granier and Pantoya [16] yields very dramatic results. Granier and Pantoya
[16] examined the ignition and combustion velocity of Al-MoO3 pressed composites as a
4
function of Al particle size. Their work showed that reducing the Al particle size from
micron to nanometer dimensions decreased the ignition delay time by two orders of
magnitude (from 10 s to 10 ms). The significantly enhanced ignition sensitivity was
attributed to the increased reactivity of the nanometer particles [17]. Pantoya and Granier
[17] showed through thermal analysis that the nano-Al and MoO3 reaction is initiated at
temperatures below the melting of Al (660 °C) and sublimation of MoO3 (770 °C) and
that the reaction takes place in the solid-solid phase. They also showed that for micron Al
reacting with MoO3, the reaction is initiated after Al melting and MoO3 sublimation such
that ignition is controlled by a liquid-gas diffusion mechanism. These results indicate that
the nano-Al particles are more reactive because they react in the solid phase and at lower
temperatures than the micron-Al particles.
1.3 Aluminum Combustion
An important part of the study of Al-MoO3 combustion is the behavior of
aluminum during the ignition process. Much work has been done related to the reactivity
of aluminum, both on the micron and nano scale, over the past 20 years [18-31]. In a
study by Popenko et al. [18], a mixture of ultrafine Al powder was combined with micron
Al powder for an examination of the bimodal Al distribution combustion behavior in air.
They analyzed the presence of bound nitrogen in the products of bimodal Al and air
combustion and found that for mixtures consisting of less than 70 % micron Al powder
the percent of bound nitrogen remained constant. The interesting finding was that the
bound nitrogen content in the combustion products of these mixtures decreases
considerably if the ultrafine Al concentration in the mixture is less than 20 % and this
5
behavior is attributed to the concurrent processes of sintering and incomplete combustion
[18].
In a study by Trunov et al. [23], differential scanning calorimetry (DSC) and
thermogravimetric analysis (TGA) were used to analyze micron sized aluminum behavior
at increased temperatures. A focus of this study was on the phase changes of the alumina
oxide layer that covers the outside of each aluminum particle. Trunov determined the
initial oxide layer to be amorphous and between two to three nanometers thick for his
samples. During the TGA analysis, a weight gain was observed between 550-650 °C.
During this temperature range, the alumina layer reportedly underwent a phase change
from amorphous to γ-phase. The γ-phase has a density that is 20% higher than the
amorphous phase and after this transition the outer surface of the aluminum particle may
not be completely surrounded by an oxide layer. Exposed aluminum would oxidize as
oxygen from the atmosphere diffuses to the particle surface, which could account for the
weight gain observed.
Other factors may also contribute to the enhanced ignition sensitivity, such as
altered absorption properties which enable nano-Al particles to more readily absorbed
energy than their micron scale counterparts. In fact, Yang et al. [32] showed that the
absorption coefficient of 30-nm Al particles is significantly greater than micron scale
particles and a strong function of particle size (see Fig. 5 of Ref. 32).
Using nanometer combined with micron scale Al particles in rocket propellant
applications has strong advantages. For example, all Al particles are pyrophoric and
therefore passivated with an unreactive oxide shell (e.g., Al2O3). As the particle surface
area to volume ratio increases the presence of Al2O3 increases and becomes a significant
6
portion of the overall mixture. Because propellant payloads can be restrictive, the
unwanted levels of an unreactive oxide that may add weight and reduce energy density
are undesirable [11, 35]. For this reason, adding small amounts of nanometer to micron-
scale Al particles may facilitate increased reactivity without the unwanted burdens of
excessive amounts of Al2O3. In a study by Dokhan et al. [36] the burning behavior of
ammonium perchlorate (AP) solid propellant with bimodal aluminum particle size
distributions was examined. They showed a significant increase in burn rate with only a
20% addition of nanometer Al. At this level, Dokhan et al. [36] showed Al combustion
takes place closer to the propellant burning surface allowing increased radiative and
conductive heat feed back that increases the temperature at the burning surface and
correspondingly increases the burn rate.
Due to the increased ignition sensitivity and burning rates of the nano-composite
thermites, a recent surge of interest is focused on developing thermites that may replace
traditional lead-based compounds in gun primers [37, 38]. Reducing the presence of
toxins such as lead in firearms will not only reduce health risks to personnel but will also
improve the environment. In particular, nanometer Al mixed with MoO3 and acetylene
black (a form of carbon) is being studied as a replacement for lead compounds [37].
This study will examine the ignition sensitivity and combustion velocity of Al and
MoO3 composites as a function of the Al particle size distribution. Mixtures are prepared
using 4 or 20 micron combined with 80 nm diameter Al particles in discrete mixture
ratios. Both powder and compressed pellet combustion was studied. The goal is to
investigate the influence of nanometer Al on the ignition sensitivity and combustion
velocity of thermites.
7
CHAPTER II
EXPERIMENTAL
2.1 Sample Preparation
Table 1 shows physical data for the Al particles and the material supplier for the
Al and MoO3. The Al particles are encapsulated within a protective Al2O3 shell. The
active Al content is the percent of Al powder that is not in the form of Al2O3. As can be
seen with the 80 nm diameter Al powder, the Al2O3 shell becomes an appreciable portion
of the total powder, causing the active Al content to reduce to only 73 % (Table 1). The
active Al content is reported by the manufacturer. The particle diameters listed in Table
1 are an average value reported by the manufacturer. Micrographs, one of which is shown
in Figure 1, reveal that the reported average particle diameter is consistent with the
particles observed using scanning electron microscopy (SEM). The average oxide shell
thickness is calculated assuming that the oxide layer is uniformly surrounding spherical
particles and all particles are of the average particle size. Based on these assumptions and
the reported active Al content and the average particle diameter, the thickness of the shell
is calculated and tabulated in Table 1.
Table 1. Reactant Particle Description
Particle % Active Al Content
Al2O3 Layer Thickness Supplier
80 nm Al 73 4.3 nm Nanotechnologies 4 um Al 91 70 nm Alfa Aesar
20 um Al 99 30 nm Sigma Aldrich MoO3 - - Technanogy/Climax
Aluminum was mixed with MoO3 in a 40/60 wt % ratio which corresponds to a
fuel rich equivalence ratio of 1.3, based on active Al content. This mixture ratio was
shown to be an optimal composition for achieving the highest combustion velocity and
shortest ignition delay time [16]. The mixtures used to make pellets were dispersed in a
hexane solution and sonicated to break up agglomerates and ensure a homogeneous
mixture. The wet solution was poured into a tray and slightly heated to allow hexane
evaporation.
Figure 1. A SEM micrograph of an 80 nm and 4 µm Al mixture.
For pellet tests, a well-mixed, dried powder was separated into 230-270 mg
quantities and cold pressed with a hydraulic press (0.5 – 1.5 MPa) and a uniaxial die. All
final pellets were 6.51 mm in diameter and 3.9 mm in length. Higher pressures (1.5
8
MPa) were needed to form the less dense nano-powder formulations to the same
dimensions of the micron-powder formulations. Theoretical maximum density (TMD)
calculations are based on the weighted average of Al, MoO3, and Al2O3 present in the
mixture. For pellets, the powders were pressed to a TMD of 50 % ( 2 g/cc) while the
pour density of powders was 5 % TMD (
≈
≈ 0.2 g/cc).
Eleven mixtures of Al/MoO3 were prepared, each with a varying distribution of
Al particle size ranging from 100 % 80 nm to 100 % 4 or 20 µm diameter. The
distribution of active aluminum by particle size for each of the samples is illustrated in
Fig. 2.
Distribution of Active Aluminum
0
10
20
30
40
50
60
70
80
90
100
1 2 3 4 5 6 7 8 9 10 11
Mixture
Perc
ent A
ctiv
e A
lum
inum
Nano AlMicron Al
Figure 2. Schematic diagram illustrating aluminum distribubimodal mixture ratios.
9
10
In all experiments, the thermite mixtures were burned in an ambient environment.
A study by Asay et al. [39] has experimentally shown that burning rates of Al/MoO3
powders were the same when in ambient and vacuum environments.
2.2 Laser Ignition of Pellets
The pellets were ignited using a 50-W CO2 laser (Universal Laser Systems Inc,
Scottsdale, AZ), power meter and associated optics. Ignition and flame propagation were
recorded using a Phantom IV (Vision Research, Wayne, NJ) high-speed camera which
captured images at 32,000 frames per second (fps). Details of this experimental apparatus
are discussed elsewhere [16].
Ignition is defined in the context of thermite combustion as the onset of a fully
sustained self-propagating reaction. There are several techniques for measuring an
ignition delay time [16]. The technique applied here is based on the “first-light” approach
in which ignition time is determined as the time lapse between sample exposure to laser
beam and detection of the first light. This may not guarantee ignition but is a commonly
used technique for experimentally determining ignition times [40]. The high-speed
camera is synchronized with the CO2 laser and detects light intensity. In this way, the
reaction light is used as the illuminating source to visualize the ignition process.
Burn rate is a measure of the burning solid surface of an energetic material and
often is used in reference to a single particle. Propellant combustion studies typically
refer to a mass burning rate which characterizes the regression rate of the combusting
solid propellant. In this thermite combustion, the flame consumes and spreads through
discrete particles packed in a highly porous matrix. The physics of flame propagation in
11
this arrangement entails flame spreading and is a strong function of the packing
arrangement of particles and the porous structure of the material. The term combustion
velocity will be used to characterize the speed of the leading edge of the reaction zone
identified by visible light emission recorded from the high-speed photographic data.
2.3 Laser Ignition of Loose Powders
A schematic of the setup used in the laser ignition of powders is shown in Figure
3. A Coherent 250-W CO2 laser (Santa Clara, CA) was used to ignite the samples. A
NaCl beam splitter (Vigo Photovoltaic, Warsaw, Poland) diverted 8% of the laser beam
to a tri-metal detector used to determine the time of the start of the laser pulse. A
Thorlabs DET210 photo-diode (Newton, NJ) with a response time of 1 nanosecond,
measured light intensity of the reaction and determined the ignition time, based on a
Tektronix (Richardson, TX) oscilloscope with a response time of 1 nanosecond. The
ignition time was defined as the time at which the photo-diode trace reached 3% of its
peak voltage.
Figure 3. Schematic diagram of the unconfined loose powder ignition sensitivity test apparatus.
2.4 Combustion Velocities of Loose Powders
The unconfined loose powder tests were completed using an apparatus
schematically shown in Figure 4. A piezo-electric starter ignited the powder mixture
(120 mg) poured into the tray. A profiling glide leveled a consistent cross section
Figure 4. Schematic diagram of the combustion velocity test apparatus for loose powders.
12
13
throughout the sample. The pour density of the powder was about 0.02 g/cc
corresponding to a TMD of 5 %. As the reaction progressed inside the tray, two holes, 1
mm in diameter, in the bottom of the tray, which were 2 cm apart, allowed light from the
reaction to be emitted into fiber-optic cables. Photo-diodes were used to convert this
light pulse to a voltage that was displayed on an oscilloscope.
The average velocity of the unconfined burn was determined from the two voltage
pulses on the oscilloscope. The time of reaction propagation between the holes was
determined by the time delay between the two voltage pulses. Due to the small size of
the hole, the voltage pulses did not occur until the reaction passed directly over the
pinholes. The average velocity was calculated by dividing the distance between the holes
by the time of reaction propagation between the holes. Previous tests have shown
through high speed photography that the reaction reaches its steady state velocity after a
very short distance, which supports the accuracy of this average velocity measurement.
2.5 Pressure Cell Tests
For the pressure cell experiments, a YAG Continuum MiniLite 100mJ laser
(Santa Clara, CA), by way of a Thorlabs fiber-optic cable, was used to ignite the samples
inside a parr-bomb type chamber. An aluminum cup contained the mixture at a constant
volume of 0.151 cm3. The aluminum cup did not react with the samples and was
weighed after each test to ensure this. A PCB Piezotronics (Depew, NY) pressure
transducer, with a response time of 1 microsecond, was used to determine maximum
pressure output for each test.
CHAPTER III
THEORY
3.1 Flame Structure
Ignition of the thermite reaction is traditionally thought to be due to a phase
change in one or more reactants. Due to the diffusive nature of the reaction, melting and
sublimation are common phase changes that allow elevated temperature fuel and
oxidizing particles to come in better contact, which then facilitates ignition of the
reaction. The Al/MoO3 reaction occurs according to the following chemical equation:
2Al + MoO3 Al2O3 + Mo. ∆Hcomb=4279cal/cm3
The phase changes of this reaction could be the melt of aluminum at 660 °C
and/or the sublimation of MoO3 at 700 °C. A simplified diagram of the flame structure is
shown in Figure 5.
Figure 5. Typical thermite flame structure diagram
14
As the reaction occurs, molten products spew away from the pellet or powder
specimen. Some of the energy from the reaction convectively heats a region of the
reactants before ignition of those particles. The thickness of the pre-heat zone is
determined by the porosity and thermal characteristics of the sample material.
3.2 Diffusion vs. Chemical Kinetics
The combustion behavior of these reactants can be explained by describing the
rate at which reactants (Al and MoO3) convert to products (Al2O3 and Mo) [41]. The
reaction can be reduced to considering a single small sphere of fuel (Al) and surrounding
oxidizer. First assume that the reaction occurs at the surface of the solid Al sphere and
the reaction is either limited by the chemical reaction rate or the oxidizer diffusion rate.
The reaction rate (ω) can be expressed as shown in Eq. (3.1) [41] assuming the following
global reaction.
32
2OAlAlCkC=ω (3.1)
2Al + MoO3 Al2O3 + Mo
In Eq. (3.1) and are the concentrations of the reacting chemical species
Al and Al
AlC32OAlC
2O3 respectively. Each concentration is raised to the power equal to the
corresponding stoichiometric coefficient. The specific reaction-rate constant, k, can be
expressed as an Arrhenius law according to Eq. (3.2) [41], where A is the pre-exponential
factor and is the activation energy of the reaction. aE
)exp(RTEAk a−
= (3.2)
15
Assuming that the spherical nano-scale particles burn on the surface as oxidizer
diffuses to the surface, the mass-burning rate ( ) of the sphere is given by Eq. (3.3) [41],
where is the mass flux at the surface, D, is the mass diffusivity and B is the mass
transfer number.
m&
"Rm&
)1ln(4"4 2 BDRmRm R +== ρππ && (3.3)
The mass flux from the particle surface would equal the fuel reaction rate. The
reaction rate is first order in oxygen concentration and second order in fuel concentration
as in Eq. (3.1).
2,," RFRoxR YkYm == ω& (3.4)
Assuming that the reaction rate occurs at the particle surface, the fuel mass
fraction ( ) in the gas phase is small compared to oxygen such that: RFY ,
0, ≈∞FY
1, =RFY
1, <<RoxY .
The mass transfer number can be expressed as
Roxox YYB ,, −= ∞ (3.5)
and
1,, << ∞oxRox YY .
The small number approximation can replace the natural log function in the mass
flux term.
)()1ln(" ,, RoxoxR YYRDB
RDm −≈+= ∞
ρρ& (3.6)
16
Solving for in Eq. (3.4) and substituting into Eq. (3.6), the mass flux at the
particle surface is given in Eq. (3.7).
RoxY ,
)(" ,
RDk
RDk
Ym oxR ρ
ρ
+= ∞& (3.7)
In diffusion controlled reactions, the reaction rates are fast compared to the
oxygen diffusion rates. Here k >> ρD/R and the mass flux is a function of the particle
radius according to Eq. (3.8).
RDYm oxR
ρ∞= ,"& (3.8)
If the reaction rates are comparable to or slower than the oxygen diffusion rate,
then the chemistry controls the reaction. In this case, k << ρD/R and the mass flux is
independent of particle radius (Eq. (3.9)).
∞= ," oxR kYm& (3.9)
For diffusion controlled reactions, the mass flux (Eq. (3.8)) and subsequent
reaction rates are inversely proportional to particle radius and directly proportional to
oxygen concentration. Eq. (3.8) suggests that smaller particles (nano-regime) will
produce higher mass flux rates than larger particles (micron regime) and thus produce
increased burn rates.
17
CHAPTER IV
RESULTS
4.1 Ignition Sensitivity
Figure 6 is a sequence of still frame images illustrating ignition and flame
propagation of a pellet composed of 60 wt % 80 nm and 40 wt % 4 µm Al particles
combined with MoO3. The sequence of images was captured with the high-speed camera
at 32,000 fps (corresponding to 31 ms intervals). Ignition occurs at the pellet center front
face and the reaction propagates radially then axially through the pellet. The flame front
is stable and planar and fully self-sustained.
s
Figure 6. Still frame images of ignian Al distribution of 60 wt % 80 nm
m1 11.762
s
111.792 ms
111.823 ms
111.855 ms
m1 11.886s
111.916 ms
111.947 mtion and flame propagation of an Al-MoO3 pellet with and 40 wt % 4 µm. The time from trigger is given.
18
Figures 7 and 8 show the ignition delay time as a function of percent nano-Al
within the pellet mixture. The standard deviation bars represent the range of
measurements for 4-6 pellets from each of the 11 mixtures, the data symbol corresponds
to the average ignition delay time. The ignition times reported here for all mixtures
containing nano-Al are consistent with previous ignition time measurements made by
Granier and Pantoya [16] for 100 % nano-Al – MoO3 mixtures.
Figure 9 shows the ignition delay time as a function of percent nano-Al for a
loose powder mixture. Both mixtures containing either 4 or 20 µm Al are shown for laser
powers set at 50 or 100 W.
0
50
100
150
200
250
0 20 40 60 80 1Percent 80nm Al
Igni
tion
Del
ay T
ime,
ms
00
Figure 7. Ignition time for pellets as a function of percent 80 nm Al content mixed with MoO3 and 4 µm Al at 50 W laser power
19
0
500
1000
1500
2000
2500
3000
0 20 40 60 80
Percent 80nm Al
Igni
tion
Del
ay T
ime,
ms
100
Figure 8. Ignition time for pellets as a function of percent 80 nm Al content mixed with MoO3 and 20 µm Al at 50 W laser power
1
10
100
1000
10000
0 20 40 60 80 1Percent 80nm Al
Ign
itio
n D
ela
y T
ime
, m
s
00
20µm and 80nm at 100W
4µm and 80nm Al at 100W
20µm and 80nm Al at 50W
4µm and 80nm Al at 50W
20
Figure 9. Ignition time for loose powders as a function of percent 80 nm Al content mixed with MoO3 and 4 or 20 µm Al at 50 or 100 W laser power.
21
4.2 Combustion Velocity
The Phantom imaging software uses a designated light intensity transition and the
elapsed time between frames to calculate the combustion velocity. The combustion
velocity as a function of 80 nm Al content is plotted in Fig. 10 for the 4 µm mixture
pellet, in Fig. 11 for the 20 µm mixture pellet and Fig. 12 for the 4 and 20 µm mixture
powders.
In Figure 11, the data for 0 and 10 % 80-nm Al content are not shown. In the case
of 0% nm Al, the pellet was exposed to the laser for a relatively long period of time
before ignition. The exposure to this heat flux volumetrically heated the pellet, igniting
the pellet under different initial thermal conditions than for the pellet mixtures containing
nano-Al. For this reason, this data was not included. For the 10 % 80 nm Al sample, the
pellet could not sustain a self propagating wave and velocity measurements could not be
made.
0
5
10
15
20
25
30
35
40
0 20 40 60 80
Percent 80nm Al
Com
bust
ion
Vel
ocity
, m/s
100
Figure 10. Combustion velocity as a function of weight percent 80nm Al content for pellets with 4 µm Al and MoO3
0
5
10
15
20
25
30
35
40
45
0 20 40 60 80
Percent 80nm Al
Com
bust
ion
Velo
city
, m/s
100
Figure 11. Combustion velocity as a function of weight percent 80nm Al content for pellets with 20 µm Al and MoO3
22
The unconfined loose powder combustion velocities for both micron-scale
mixtures are plotted versus the percent nano Al content in Fig. 12. The piezo-electric
starter was not able to ignite mixtures with 0% or 10% nano Al content, therefore no data
are shown for these mixtures.
0
50
100
150
200
250
300
350
400
450
0 20 40 60 80 100
Percent 80nm Al
Co
mb
ust
ion
Ve
loci
ty,
m/
s
20µm and 80nm Al4µm and 80nm Al
Figure 12. Combustion velocity as a function of weight percent 80nm Al content for loose powder with 4 or 20 µm Al and MoO3.
23
24
CHAPTER V
DISCUSSION
5.1 Ignition Sensitivity
The increased ignition sensitivity of mixtures with nanometer scale Al may be
explained by the increased reactivity of nano-sized metal particles. Pantoya and Granier
[16] showed through thermal analysis on nano and micron Al combined with MoO3
mixtures that the initiation mechanisms for the two particle length scales are very
different. They showed that for micron composites, Al oxidation occurs after Al melting
(660 °C) and MoO3 sublimation (770 °C) and in the solid-solid phase [16]. Ignition
temperatures corresponding to micron composites were measured at roughly 1000 °C and
the diffusion mechanism is in the liquid-gas state [16]. Figures 7-9 show that only 10 wt
% nano-Al content is required to reduce ignition delay times by two orders of magnitude.
Enough energy is generated from the localized reaction between nano-Al and
MoO3 to ignite neighboring particles. However, if the micron scale particles are too large
(20 µm) and only 10 % nano Al is within the matrix, enough heat is lost to the larger
scale particles to prevent a self-sustained reaction. In this case, shortly after a localized
ignition spot is formed, the reaction is extinguished. It is noted that Popenko et al. [18]
similarly observed that if the nano-Al concentration in a mixture with micron scale Al
powder is less than 10 %, combustion is difficult to initiate.
The effective thermal conductivity of the composite may also influence
ignition sensitivity. The effective thermal conductivity is a function of the thermal
conductivity values of each component within the matrix, the volume fraction, and the
25
distribution of the matrix and dispersed phases(s). Hasselman and Donaldson [42]
theoretically investigated the role of material distribution in the effective thermal
conductivity of composites and found that the presence of an interfacial thermal barrier
could have a significant effect on composite thermal conductivity. In this system, the
Al2O3 shell encapsulating nano-Al particles could represent an interfacial thermal barrier.
As localized Al oxidation reactions occur, the Al2O3 may provide enough thermal
insulation to facilitate localized energy build up. When energy is constrained, heat losses
associated with thermal diffusion away from the localized reaction are reduced. This
reduction in heat losses effectively causes localized hot spots thermally insulated from
micron-scale Al particles.
Another factor that could contribute to the increased ignition sensitivity
associated with nano-Al is the physical change of the alumina layer during heating. The
oxide layer of the nano-Al particles is less than 5 nm thick and amorphous (Table 1).
Upon heating, Trunov and co-workers [23] showed that the amorphous alumina shell
undergoes a phase transition just below 500°C to form γ-Al2O3. This phase change
corresponds to an increase in density by about 20 % (from 3050 to 3660 kg/m3) [23].
This density increase will cause the aluminum core to become exposed to air within void
pockets and surrounding MoO3 particles. Because this phase transition occurs at
temperatures corresponding to the ignition temperature of the nanocomposite, exposure
of the solid-Al core may be a rate-determining step in the solid-solid diffusion
mechanism of the nano-Al and MoO3 reaction. This step may not be as critical in the
ignition of micron-composites because the aluminum surface particles of the aluminum
core of a micron-Al particle make up a much smaller portion of the total aluminum
26
content. When considering the same amount of active aluminum for the nano-Al and
micron-Al particles, nano-Al will have much more aluminum core surface area that can
be exposed during the phase transition of the Al2O3 shell since there are so many more
particles for the same amount of mass. The energy released from the nano-Al during this
step would be much larger than the energy release of the micron-Al, which could greatly
increase ignition sensitivity of these nanoparticles.
5.2 Combustion Velocity
The combustion wave speeds shown in Figs. 10 and 11 increase from roughly 1 to
40 m/s as nanometer Al content increases. The most interesting behavior is observed in
Fig. 11 in which a sharp transition from relatively slow to fast flame propagation occurs
between 50 and 70 % nanometer Al content. A similar trend was previously observed in
porous explosive charges and attributed to a transition from normal to convective burning
[43]. When convective burning takes precedence over thermal conduction and radiation,
energy and mass transfer in the burning zone are driven by gas jets that penetrate into the
pores of the energetic material. Bobolev et al. [43] explain that in some cases penetration
of the combustion into the pores is followed by the establishment of a regime of
stationary convective burning whose rate substantially exceeds the normal burning rate.
In an effort to identify this burning regime, an analysis of porous energetic material
combustion under constant pressure conditions has lead to a stability criterion known as
the Andreev number, An (Eq. (5.1)) [43]. This non-dimensional parameter is similar to
the Peclet number except tailored for reacting flow through porous media.
An=g
phb
kcUdρ
=const. (5.1)
In this equation, ρb is the bulk density of the composite, U is the measured combustion
wave speed, dh is the hydraulic pore diameter, cp is the heat capacity of the composite and
kg is the thermal conductivity of the gas. If this value exceeds a certain constant,
combustion penetrates into the porous structure and convective mechanisms dominate
flame propagation.
In making this calculation it is necessary to estimate the characteristic pore size
which is termed the hydraulic pore diameter, dh, given by Eq. (5.2) [44].
)1(4
εε−
=o
h Ad (5.2)
In this equation ε is the bulk void volume, Ao is the specific surface area based on the
solid volume and is calculated as the solid surface area divided by the solid volume
(As/Vs), and (1-ε) is the solid volume fraction [44].
Table 2 shows the calculated values of the physical and thermal parameter for
both the 4 and 20 micron containing bimodal mixtures. These calculations are based on
the assumption of kg = 0.137 W/mK corresponding to air at an average temperature of
2000 K. For each estimate, the An number is significantly less than 1.0. Bobolev et al.
[43] calculated critical An numbers between 3 and 10 and generalized that if the An > 6.0
combustion will penetrate into the pores and convective mechanisms will play the
primary role in accelerating the flame front.
27
Table 2. Calculated thermal and physical parameters for all Al-MoO3 composites
0 2272 885.4 0.165 1.2678E-03 1.98E-08 5.27E-14 3.24E-0710 2270 884.4 0.982 7.5389E-03 1.97E-08 8.83E-15 1.78E-0720 2268 883.4 1.798 1.3810E-02 1.96E-08 4.80E-15 2.83E-0730 2266 882.5 2.615 2.0081E-02 1.95E-08 3.29E-15 1.92E-0740 2264 881.5 3.431 2.6352E-02 1.94E-08 2.50E-15 1.33E-0750 2262 880.5 4.248 3.2623E-02 1.93E-08 2.01E-15 2.85E-0760 2260 879.5 5.064 3.8894E-02 1.92E-08 1.68E-15 3.19E-0770 2258 878.5 5.881 4.5165E-02 1.91E-08 1.44E-15 3.70E-0780 2256 877.5 6.698 5.1437E-02 1.90E-08 1.26E-15 3.93E-0790 2253 876.5 7.514 5.7708E-02 1.89E-08 1.11E-15 3.56E-07100 2251 875.5 8.331 6.3979E-02 1.87E-08 9.99E-16 5.44E-07
0 2278 889.8 0.033 2.5289E-04 2.01E-08 2.67E-1310 2276 888.4 0.863 6.6255E-03 2.00E-08 1.02E-1420 2274 887.0 1.692 1.2998E-02 1.99E-08 5.16E-15 7.60E-0830 2272 885.5 2.522 1.9371E-02 1.98E-08 3.45E-15 6.89E-0840 2269 884.1 3.352 2.5743E-02 1.96E-08 2.58E-15 7.68E-0850 2267 882.7 4.182 3.2116E-02 1.95E-08 2.06E-15 1.53E-0760 2264 881.2 5.012 3.8488E-02 1.94E-08 1.71E-15 3.43E-0770 2261 879.8 5.841 4.4861E-02 1.92E-08 1.46E-15 6.26E-0780 2258 878.4 6.671 5.1234E-02 1.91E-08 1.27E-15 6.33E-0790 2254 876.9 7.501 5.7606E-02 1.89E-08 1.12E-15 5.90E-07100 2251 875.5 8.331 6.3979E-02 1.87E-08 9.99E-16 4.37E-07
Hydraulic Pore Diameter (m)
Andreev Numberε/(1−ε)Αο (1/m)% nano Al
contentTotal Surface
Area (m2)ρ (kg/m3) Cp (J/kgK)
For 80 nm Al mixed with 20 micron Al particles and MoO3
For 80 nm Al mixed with 4 micron Al particles and MoO3
Αο (1/m) ε/(1−ε)Hydraulic Pore Diameter (m)
Andreev Number
% nano Al content ρ (kg/m3) Cp (J/kgK)
Total Surface Area (m2)
The above analysis suggests that the transition from low to high combustion wave speeds
may not result from a shift in the flame propagation mechanism. A better explanation for
this behavior may be related to the ignition insensitivity of the micron-scale Al particles.
Because the micron-scale particles require more time and higher temperatures to ignite,
the nano-Al reactions may proceed too quickly to allow the micron-Al particles to
participate in the reaction. In this way, reactants containing large amounts of micron-Al
particles may experience significant incomplete combustion which would result in
reduced combustion velocities.
For mixtures with mostly nano-Al particles (70-90% nano Al), the micron-Al
particles make up a significantly smaller portion of the volume and impede the nano-Al/ 28
MoO3 reaction much less. Channels of nano-Al and MoO3 exist throughout the powder
or pellet, allowing the reaction to propagate without micron Al obstructions hindering the
velocity. These channels allow the measured velocity to be the same as a mixture of pure
nano-Al and MoO3 but may still result in incomplete reactions between micron Al and
MoO3.
To test the theory that incomplete combustion is responsible for the reduced
velocities in Figures 10-12, a series of pressure measurements were performed using the
pressure cell described previously. Figure 13 shows the peak pressure of both the 4 and
20 µm Al loose powder mixtures as a function of percent nano-Al content. The pressure
is displayed on a per unit mass basis because a slight pour density change occurred as the
size distribution of the aluminum particles changed. The constant volume sample cup
0
5
10
15
20
25
30
35
0 20 40 60 80 1
Percent 80nm Al Content
Pres
sure
Out
put,
kPa/
mg
00
20µm and 80nm Al4µm and 80nm AlCalculated Pressure
Figure 13. Pressure output as a function of percent nano-Al content for loose powder Al-MoO3 mixtures with 4µm and 20µm Al.
29
holds more material if composed of the more-dense micron particles. The sample amount
slightly increased as a result, but, more importantly, the volume of the pressure vessel
remained constant. No data points exist for the 0-20% nano-Al content due to the
difficulty of igniting the mixture with the YAG laser.
The maximum peak pressure of the powder samples increases as the amount
of nano-Al in each mixture increases (Figure 13). The pressure increase corresponds
linearly to the increase in nano-Al content. This trend suggests that the micron-Al
particles do not contribute to raising the peak pressure. For example, assuming a
complete reaction between all Al particles and MoO3, the maximum peak pressure for the
mixture can be calculated assuming the peak pressure results from the rapid expansion of
high temperature gasses present in the parr-bomb chamber under roughly adiabatic flame
temperature conditions. Equation 4.3 is the ideal gas law, which assumes that the gas
behavior at the flame temperature is ideal.
VmRTP = (4.3)
In this equation, P is the peak pressure, T is evaluated at the adiabatic flame
temperature of the mixture 3200 °C [1], which is sufficiently high enough to consider an
ideal gas assumption. The gas constant (R) for air at 3200 °C is used in conjunction with
the volume (V) of the combustion chamber, 9.26cm3. The mass (m) of gas present in the
chamber is estimated as 8.75 mg based on [1] which states that 25% of the products of
this reaction are gaseous. With these values in Eq. (4.1), the maximum peak pressure for
a complete reaction is 25.2 kPa/mg, and is shown in Fig. 13 as the dashed line. The
measured peak pressures approach 25.2 kPa/mg only when a significant percent of nano-
30
Al is included in the mixture, indicating that the micron-Al particles may not be
contributing to the overall reaction enough to raise the peak pressure to the theoretical
value.
Increased levels of nano-Al may also enhance the radiant heat transport
through the mixture. Yang et al. [32] showed nano-scale Al particles absorb more energy
compared to their micron scale counterparts. The increased levels of absorbed energy
may generate a more intense radiation field and elevate the temperature of the preheat
zone. Higher preheat temperatures will induce increased combustion velocities. If too few
nano-particles are present in the composite, the thickness and intensity of this preheat
zone will diminish and result in comparably slower combustion velocities.
Still frame images and light traces captured during powder ignition experiments
also complement the theory that micron scale particles are either not participating in the
reaction or react much slower than the reaction between nano-Al and MoO3. The six
frames of Figures 14-16 show the combustion of three different powder mixtures. The
frames of Figure 14 show the reaction of 100% 80nm Al and MoO3.
Figure 14. Combustion of 100% 80nm Al and MoO3 loose powder mixture
31
The reaction appears to consume all particles, which rise into the air and convectively
cool as smoke. More importantly, all of the particles seem to react at the same time,
although a few get forced out onto the sample platform as seen in the third frame. This
whole reaction is outside the range of the photodiode in as little as 9 ms..
A different looking reaction appears in Figure 15, which is a mixture of 10%
80nm Al and 90% 20 µm Al with MoO3. The reaction of the nano Al and MoO3 can be
seen in the first two frames with a convective plume similar to Figure 14. Many other
particles can be seen radiating in these frames. The micron Al particles may only begin
reacting during the nano-Al reaction with MoO3 and forced into the air while still
radiating and reacting. These radiating particles appear to burn much slower and are
Figure 15. Combustion of loose powder Al and MoO3 mixture with 10% 80nm Al and 90% 20µm Al
32
Figure 16. Combustion of 100% 4µm Al and MoO3 loose powder mixture.
apparent even after the nano-Al plume has disappeared. The time duration of these six
frames is about 13 times the time duration of the frames of Figure 14; therefore, it can be
seen that the micron particles react more slowly than the nano Al with MoO3. This
finding is consistent with the peak pressure measurements which indicate that the micron
particles do not contribute to increasing the measured peak pressure.
The frames of Figure 16 show a reaction with 4 µm Al and MoO3. This reaction
also occurs slowly relative to the reaction in Figure 14, about 2.5 times as slow. The
reaction spreads from the left to the right across the sample cup in the first three frames
and sends a bright plume of radiating particles into the air. Radiating particles can still be
seen floating in the air in the final two frames. It is apparent that the reaction of micron
Al particles, even as small as 4 µm, with MoO3 occurs much more slowly than the nano
Al reaction.
Products of a pellet reaction with 50% 80 nm Al and 50% 20 µm Al mixed with
MoO3 are shown in Figure 17. These micron particles were removed from the test
apparatus after combustion. The surfaces of these particles are of both smooth and rough
nature. The smooth sections are typically indicative of unreacted Al, as seen in Figure 1.
33
The rough sections, however, indicate aluminum that has reacted with oxygen. These
SEMs show that micron Al particles in this mixture are not completely reacting, which
supports the still frame images of Figures 15 and 16.
A. B.
C.
Figure 17. SEM images of products of 50% 80 nm Al and 50% 20 µm Al pellet mixtures with MoO3
34
35
The results obtained for bimodal mixtures in a thermite composite are similar to
results of bimodal Al mixtures reacting with air made by Popenko [18]. They suggest
incomplete combustion for nano-Al concentrations less than 20 % [18]. This finding is
consistent with the slower reacting (or incomplete combustion) micron Al particles
observed in this study. Also, Dokhan [36] showed a plateau in burning performance for
70 % nano-Al concentration in bimodal Al mixtures added to AP. Our work suggests 70
% nano-Al concentration will optimize the combustion wave speed in thermites.
The results presented here may have implications towards the use and
handling of thermites. Nano-scale Al particles can be costly, significantly more so than
micron scale particles. It may be advantageous to use mixtures of nano and micron scale
material for large scale formulations. This work shows that ignition sensitivity is
heightened by small additions of nano-Al to the mixture. However, what is gained in
ignition sensitivity is sacrificed in performance through reduced velocities that result
from incomplete (or significantly slower) micron-Al reactions.
36
CHAPTER VI
CONCLUSIONS
The ignition delay time and combustion wave speed of thermite composites
composed of Al and MoO3 were examined as a function of Al particle size distribution.
Bimodal Al size distributions consisting of 80 nm combined with either 4 or 20 µm
particles showed increased sensitivity to ignition with only a 10 % concentration of nano-
scale particles. The increased ignition sensitivity is explained by the solid-solid phase
ignition mechanism associated with nano-Al particles compared with the liquid-gas phase
mechanism associated with micron-Al particles. The nano-Al particles are more reactive
and stimulate ignition two orders of magnitude faster than with purely micron-Al
containing mixtures. Although the energy generated from the localized reaction between
nano-scale Al and oxidizer is enough to initiate the reaction, in some cases the reaction is
not self-sustaining. For example, with 10 % nano-Al particles combined with 20 micron
particles the initiated reaction is quickly quenched as heat is lost to the surrounding media
too quickly to permit a sustained reaction. This study indicates that at least 20 % nano-Al
particles are required to ensure reduced ignition delay times and are on the same order as
with mixtures containing 100 % nano-Al.
The combustion velocity increases linearly with nano-Al content from roughly 1
to 40 m/s. This is consistent with prior investigations on bimodal Al size distributions
concerning Al with air and Al as a propellant additive. Slow (and incomplete)
combustion of micron Al particles leads to the decrease of the combustion wave speed for
mixtures in which the micron Al particles make up a large part of the volume. For
mixtures with small amounts of micron Al, the reaction speed is significantly increased
since the nano Al can react with MoO3 in channels that exist around the micron particles.
37
CHAPTER VII
IMPLICATIONS AND FUTURE RESEARCH
Research in nanotechnology has advanced in the past decade. Within the field of
nano-energetic materials, research is still needed to help characterize the materials for
many combustion and thermal properties. Safety, longevity, and reliability are all factors
that are important in the use and predictability of these materials, and further research in
these areas will expediate the use of nano-energetics in many applications.
7.1 Nano-scale Aluminum Properties
Nano-aluminum obviously has striking combustion property differences from
micron-aluminum, as seen in this work, but a potential for significant differences in its
physical properties may also exist. One of these physical properties that has not been
extensively studied is radiation absorbance. Nano-aluminum appears a very dark gray
color, while micron-aluminum is light gray. This color difference shows optical
properties of aluminum change in the visible light spectrum with decreased particle size,
possibly due to dependent scattering. Groundwork has been laid on the potential for this
property difference [32, 33, 34], but the range of wavelengths researched is not broad.
Integrating spheres have been used for gaining reflectance information of
aluminum surfaces [45] and could be very useful for the evaluation of powder aluminum
samples. Lindseth et al. [45] experimentally determined that the rolling of aluminum
surfaces can reduce the reflectivity of the surface by over 10%. Thermites and other
38
nano-aluminum mixtures are often ignited using laser irradiation, and the absorbance,
transmittance and reflectance properties of nano-aluminum could explain which
wavelengths are most effective in this ignition process.
An important safety consideration of many materials, especially energetic
materials, is the risk of dust explosions, to which even common grains are susceptible.
The risk of micron-sized aluminum in regard to random dust explosions has been heavily
documented, but due to its recent development, nano-aluminum risks in this area have not
been studied. The enhanced ignition capabilities of nano-aluminum compared to micron-
aluminum could pose a much higher threat to explosion and must be assessed to sustain
proper usage procedures. Nano-sized aluminum is much less dense than micron
aluminum in its powder form and can more easily be dispersed in air due to its increased
surface area. The addition of nano-Al to explosives, for increased energy output, poses
another significant risk that should be assessed. Kwok et al. [30] have shown that the
addition of nano-scale aluminum to ammonium perchlorate decreases the minimum
electro-static discharge ignition energy from greater than 156 mJ to 6 mJ. In order to
safely handle nano-Al, more research is being done to quantify the risk of nano-Al dust
explosions and the proper precautionary measures for prevention.
7.2 Mixing Techniques
The enhanced performance of nano-composites over their micron counterparts is
largely due to the superior mixing distribution of the fuel and oxidizing particles. In
order to harness the full potential of these materials, techniques to produce optimum
39
mixing of these particles are necessary. Also, inconsistent performance of these materials
that could hinder their reliability could be reduced through research in this area. Current
mixing techniques utilize sonication of alcohols and other non-reactive solvents to break
up particle agglomerations and create a homogeneous mixture of fuel and oxidizing
particles. Research regarding mixing behaviors has begun [46], and sol-gel chemistry
solutions through nano-structuring are also being investigated [47-49]. Mixing
techniques will continue to be an important factor in the development and production of
nano-energetic applications in the future.
7.3 Preliminary Molybdenum Trioxide Study
After completing the bimodal aluminum distribution study, a preliminary study
was performed on the combustion effects related to environmental alterations of
molybdenum trioxide. During lab operations, a color change was noticed in molybdenum
trioxide nanoparticles that were exposed to light and the effects of the color change on
combustion speeds were unknown. This drew the attention to what environmental
conditions caused this change and what other environmental conditions could cause the
material to react differently.
Metal oxides like molybdenum oxide have been known for over 40 years to
exhibit photochromic properties [50-57]. Ultraviolet (UV) light has been shown to cause
several chemical, mechanical, and electrical properties of metal oxides to change.
Indium oxide thin films have been shown to become much more electrically conductive
through exposure to UV radiation [55]. When in an oxygen-present environment,
40
amorphous zinc oxide thin films will crystallize due to UV irradiation [56].
Photodarkening in these same zinc oxide films has been attributed to photoreduction
effects. Li has shown similar UV-induced color formation and photoreduction
observations in tungsten oxide nanoparticles, which are claimed to describe a change of
the oxide from WO3 to W2O5 [58].
Molybdenum oxide also exhibits photochromic properties. S. K. Deb [57] reports
the formation of “color centers” which result in a significant increase in optical density
due to UV irradiation. He describes this color change as being due to an increase in
oxygen vacancies in the lattice structure. He also describes a thermal bleaching process,
which entails heating the sample to 300 ºC in the presence of oxygen. The bleached
molybdenum oxide particles are rid of these color centers and do not form new color
centers in the presence of UV irradiation. In the past ten years, new technologies have
allowed the formation of nanoparticle molybdenum oxide (MoO3), which seem to have a
stronger photochromic properties. Absorbance measurements by Li have shown
nanoparticle MoO3 to be 15 times more absorbent of UV irradiation than their bulk MoO3
counterparts.
Raman spectroscopy has also been used to characterize MoO3 [59-60]. According
to S. H. Lee [59], heat treatment of amorphous thin films, in a manner similar to the
thermal bleaching described previously, causes alpha MoO3 crystal formation to occur.
Raman spectroscopy experiments from Lee shown sharp peak formation previously not
seen with the amorphous samples.
41
More information in this area of research is needed to describe variability and
inconsistency seen in performance parameters, such as combustion rate and ignition
sensitivity. Plantier has shown standard deviations of Al/Fe2O3 formulations to be as
high as 15% of the mean combustion velocity [49]. Granier has shown ignition time and
combustion speed standard deviations in Al/MoO3 formulations of greater than 20% of
the mean values [61].
Although the ability of MoO3 optical properties to change through exposure to
UV irradiation is known, little is known about how the combustion properties are affected
by similar environmental factors. A change in combustion characteristics could lead to a
greater knowledge of causes of variations in the performance of nano-composites. This
information would lead to the development of more consistent nano-composites and
expediate the use of environmentally responsible primer formulations. Therefore, the
scope of this preliminary study was to evaluate how the combustion behavior of nano-
composite Al/MoO3 formulations is affected by MoO3 particles that have been altered
through exposure to fluorescent light, UV irradiation and high relative humidity levels,
common environmental factors which may change the combustion performance of these
materials.
The aluminum used in this study was 120 nm, and the same molybdenum oxide
product from Climax was used as in the bimodal distribution study. This molybdenum
oxide was tested in its original form as received from Climax and after a heat treatment of
400°C. SEMs of the untreated and heat treated MoO3 are shown in Figure 18.
42
a. b.
Figure 18. SEM micrographs of a. Untreated MoO3 and b. Heat treated MoO3
Powders were prepared in a process similar to the preparation process mentioned
previously and were tested in an open channel setup with spark ignition. The
molybdenum oxide, both untreated and heat treated, was exposed to three different
environmental conditions for up to 4 days before being mixed with the aluminum: UV
light exposure, fluorescent light exposure, and a 99% relative humidity environment.
Figures 19-21 show the combustion velocities as a function of exposure to
fluorescent light, UV light, and high humidity levels.
43
0
50
100
150
200
250
300
350
400
450
500
0 20 40 60 80 10Exposure to Fluorescent Light (hrs)
Bur
n Ve
loci
ty (m
/s)
0
Untreated Heat Treated
Figure 19. Burn rate of Al/MoO3 as a function of fluorescent light exposure
0
50
100
150
200
250
300
350
400
450
0 20 40 60 80 10
Exposure to UV (hrs)
Bur
n Ve
loci
ty (m
/s)
0
UntreatedHeat Treated
Figure 20. Burn rate of Al/MoO3 as a function of UV light exposure
44
0
50
100
150
200
250
300
350
400
0 20 40 60 80 10
Exposure to 99% Relative Humidity (hrs)
Bur
n Ve
loci
ty (m
/s)
0
UntreatedHeat Treated
Figure 21. Burn rate of Al/MoO3 as a function of 99% relative humidity
The untreated MoO3 powder changed color from yellow to blue after both
fluorescent and UV light exposure, which has been attributed to a desporption of oxygen
from the molecules [57]. However, these types of exposure had little effect on the burn
rate of the mixture. The white heat treated MoO3 did not change color after both light
exposures, due to the differing crystal phase structure shown in Figure 18. Like the
untreated MoO3, light exposure had little effect on the burn rate.
Although light had little effect on the combustion performance for both MoO3
samples, exposure to high humidity levels drew a clear difference between the samples.
The untreated MoO3 sample burned at speeds less than a meter per second after just one
45
day of exposure, which was probably due to water molecules that had absorbed onto the
MoO3 particles. The heat treated MoO3 sample showed no change after two days
exposure and a moderate decrease thereafter. The heat treated MoO3 has not been
thoroughly studied and could provide an alternative for untreated MoO3 in harsh
environmental applications.
Based on this preliminary study, it seems environmental factors can play a
significant role in the performance of thermites. Increased humidity levels can reduce the
combustion performance of thermites very quickly, an light exposure has the potential to
decrease the performance over a longer period of time. In order to utilize these materials
in ordinance applications, environmental factors must be addressed to accurately predict
how these materials will perform. Future research could provide a more definite
correlation of the effects of these and other environmental factors on thermite materials.
46
REFERENCES
1. Wang, L.L., Munir, Z.A., and Maximov, Y.M. Journal of Materials Science 28 (1993) 3693-3708.
2. Fischer, S.H. and Grubelich, M.C. “Theoretical Energy Release of Thermites, Intermetallics, and Combustible Metals.” Proceedings of the 24th International Pyrotechnics Seminar, Monterey, CA, July 1998.
3. Varma, A. “Form from Fire.” Scientific American. August (2000) 58-61.
4. Mukasyan, A.S., Rogachev, A.S., Varma, A. Chemical Engineering Science 54 (1999) 3357-3367.
5. Varma, A., Rogachev, A.S., Mukasyan, A.S., and Hwang, S. Proc. Natl. Acad. Sci. USA 95 (1998) 11053-11058.
6. Hunt, E.M. and Pantoya, M.L. “Nanocomposite Reactions in the Self-Propagating High-Temperature Synthesis of Nickel Alumnide Systems” Proceedings of the Third Joint Meeting of the U.S. Sections of The Combustion Institute (2003).
7. Shimizu, A. and Saitou, J. Solid State Ionics 38 (1990) 261-269.
8. Brown, M.E., Taylor, S.J., and Tribelhorn, M.J. Propellants, Explosives, Pyrotechnics 23 (1998) 320-327.
9. Tomasi, R., and Munir, Z.A. J. Am. Ceram. Soc. 82 [8] (1999) 1985-1992.
10. Drennan, R.L. and Brown, M.E., Thermochimica Acta 108 (1992) 223-246.
11. Pesiri, D., Aumann, C.E., Bilger, L., Booth, D., Carpenter, R.D., Dye, R., O’Neill, E., Shelton, D. and Walter, K.C. “Industrial Scale Nano-Aluminum Powder Manufacturing” To appear in Journal of Pyrotechnics (2004)
47
12. Peterson, R.P., Mang., J.T., Son, S.F., Asay, B.W., Fletcher, M.F., Hjelm, R.P. and Roemer, E.L. “Microstructural Characterization of Energetic Materials” Proceedings of the 29th International Pyrotechnics Seminar (2002).
13. Mang, J.T., Hjelm, R.P., Peterson, P.D., Jorgensen, B.S. and Son, S.F. “Characterization of Components of Nano-Energetics by Small-Angle Scattering Techniques.” Proceedings of the 31st International Pyrotechnics Seminar (2004).
14. Son, S.F., Asay, B.W., Busse, J.R., Jorgensen, B.S., Bockmon, B. and Pantoya, M. “Reaction Propagation Physics of Al/MoO3 Nanocomposite Thermites” Proceedings of the 28th International Pyrotechnics Symposium (2001).
15. Bockmon, B.S., Pantoya, M.L., Son, S.F. and Asay, B.W., “Burn Rate Measurements in Nanocomposite Thermites,” Proceedings of the Americal Institute of Aeronautics and Astronautics Aerospace Sciences Meeting, Paper No. AIAA-2003-0241 (2003).
16. Granier, J. J. and Pantoya, M. L., Combustion and Flame 138 (2004) 373-383.
17. Pantoya, M. L. and Granier, J. J., to appear in Propellants Explosives Pyrotechnics (2004).
18. Popenko, E.M., Il’in, A. P., Gromov, A. M., Kondratyuk, S. K., Surgin, V. A., and Gromov, A. A., Combustion, Explosion, and Shock Waves 38 [2] (2002) 157-162 .
19. Il’in, A.P., Gromov, A.A. and Yablunovskii, G.V., Combustion, Explosion, and Shock Waves, 37 [4] (2001) 418-422.
20. Il’in, A.P., Yablunovskii, G.V., Gromov, A.A., Popenko, E.M. and Bychin, N.V. Combustion, Explosion, and Shock Waves 35 [6] (1999) 656-659.
21. Kwon, Y., Gromov, A.A., Ilyin, A.P., Popenko, E.M., Rim, G. Combustion and Flame 133 (2003) 385-391.
48
22. Il’in, A.P., Popenko, E.M., Gromov, A.A., Shamina, Y.Y. and Tikhonov, D.V., Combustion, Explosion, and Shock Waves 38 [6] (2002) 665-669.
23. Trunov M.A., Schoenitz M., Zhu X. and Dreizin E.L., “Effect of Polymorphic Phase Transformations in Al2O3 Film on Oxidation Kinetics of Aluminum Powders,” submitted to Combustion and Flame, June 2004.
24. Dreizin, E.L., Progress in Energy and Combustion Science 26 (2000) 57-78.
25. Dreizin, E.L. Combustion, Explosion, and Shock Waves 39 [6] (2003) 681-693.
26. Dreizin, E.L. Combustion and Flame 105 (1996) 541-556.
27. Driezin, E.L., Combustion and Flame 117 (1999) 841-850.
28. Jones, D.E.G., Brousseau, P., Fouchard, R.C., Turcotte, A.M. and Kwok, Q.S.M. Journal of Thermal Analysis and Calorimetry 61 (2000) 805-818.
29. Ilyin, A., Gromov, A., et al. Propellants, Explosives, Pyrotechnics 27 (2002) 361-364.
30. Kwok, Q.S.M., Fouchard, R.C., Turcotte, A., Lightfoot, P.D., Bowes, R., and Jones, D.E.G., Propellants, Explosives, Pyrotechnics 27 (2002) 229-240.
31. Sandstrom, M.M., Oschwald, D.M. and Son, S.F. “Laser Ignition of Aluminum Nanoparticles in Air.” Proceedings of the 31st International Pyrotechnics Seminar (2004).
32. Yang, Y., Wang, S., Sun, Z., and Dlott, D. D, J. Appl. Phys. 95 [7] (2004).
33. Yang, Y., Sun, Z., Wang., S., Dlott, D.D. J. Phys. Chem. B 107 (2003) 4485-4493.
34. Wang, S., Yang, Y., Sun, Z. and Dlott, D.D. Chemical Physics Letters 368 (2003) 189-194.
49
35. Kramer, M. “Nanotechnology Aids the Advanced Energetics Program” AFRL Technology Horizons, March 2003.
36. Dokhan, A., Price, E. W., Seitzman, J. M., Sigman, R. K., “The Effects of Bimodal Aluminum with Ultrafine Aluminum on the Burning Rates of Solid Propellants,” Proceedings of the Combustion Institute, 29 [2] (2002) 2939-2945.
37. Giles, J., Nature 427 (2004) 580-581.
38. Naud, D.L., Hiskey, M.A., Son, S.F., Busse, J.R. and Kosanke, K. Journal of Pyrotechnics 17 (2003)
39. Asay, B. Son, S., Busse, J., and Oschwald, D. “Observations on the Mechanism of Reaction Propagation in Metastable Intermolecular Composites.” Proceedings of the APS Schock Physics Meeting, Portland OR, July 2003
40. Ostmark, H. and Roman, N., J. Appl. Phys. 73 [4] (1993) 1993-2003.
41. Turns, S.R., An Introduction to Combustion: Concepts and Applications, McGraw-Hill Companies, Inc. New York, 2000
42. Hasselman, D. P. H and Donaldson, K.Y., “The Role of Particle Size in the Effective Thermal Conductivity of Composites with an Interfacial Thermal Barrier,” Journal of Wide Bandgap Materials, 7 [4] (2000).
43. Bobolev, V.K., Margolin, A.D., and Chuiko, S.V., “ Stability of Normal Burning of Porous Systems at Constant Pressure,” Combustion, Explosion and Shock Waves, 2 [4] (1966) 24-32.
44. Kaviany, M., Principles of Heat Transfer In Porous Media, Springer-Verlag, New York (1991)
45. Lindseth, I., Bardal, A., and Spooren, R. Optics and Lasers in Engineering 32 (2000) 419-435.
46. Lee, C.S., Lee, J.S. and Oh, S.T., Materials Letters 57 (2003) 2643-2646.
50
47. Tillotson, T.M., Gash, A.E., Simpson, R.L., Hrubesh, L.W., Satcher, Jr., J.H. and Poco, J.F. Journal of Non-Crystalline Solids, 285 (2001).
48. Gash, A.E., Tillotson, T.M., Satcher, J.H., Poco, J.F., Hrubesh, L.W., Simpson, R.L. Chemistry of Materials, 13 [3] (2001) 999 -1007.
49. Plantier, K., Pantoya, M.L., and Gash. “Combustion Behavior of Sol-Gel Synthesized Nanocomposite Thermites,” Proceedings of the Third Joint Meeting of the U.S. Sections of The Combustion Institute, March 2003
50. Fujita, Y. and Kwan T. Bull. Chem. Soc. Jpn. 31 [380] (1958)
51. Zelaya-Angel, O., Menezes, C. and Sanchez-Sinencio, F. J. Appl. Phys. 51 [11] (1980)
52. Yao, J. N., Hashimoto, J. N. and Fujishima, A. Nature 355 [624] (1992)
53. Reyes-Betanzo, C., Herrera-Perez, J. L., Cocoletzi, G. H., Zelaya-Angel, O. J. Appl. Phys. 88 [1] (2000)
54. Itoh, M., Hayakawa, K., Oishi, S. J. Phys.: Condens. Matter 13 [6853] (2001)
55. Bender, M., Natsarakis, N., Gagaoudakis, E., Hourdakis, E., Douloufakis, E., Cimalla, V., and Kiriakidis, G. J. Appl. Phys. 90 [10] (2001)
56. Asakuma, N., Hirashima, H., Imai, H., J. Appl. Physics. 92 [10] (2002)
57. Deb, S. K. and Chopoorian, J. A. J. Appl. Physics. 37 [13] (1966)
58. Li, S., Germanenko, I., and El-Shall, M. S. J. Cluster Science, 10 [4] (1999)
59. Lee, S., Seong, M. J., Tracy, C. E., Mascarenhas, A., Pitts, J. R., and Deb, S. Solid State Ionics. 147 [129] (2002)
51
60. Zhao, Y., Jingguo, L., Zhou, Y., Zhang, Z., Xu, Y., Naramoto, H., and Yamamoto, S. J. Phys.: Condens. Matter. 15 [L547] (2003)
61. Granier, J., and Pantoya, M., Laser ignition study of medium density nanocomposite thermites. Submitted to Combustion and Flame (May 2003)
52
PERMISSION TO COPY
In presenting this thesis in partial fulfillment of the requirements for a master’s
degree at Texas Tech University or Texas Tech University Health Sciences Center, I
agree that the Library and my major department shall make it freely available for research
purposes. Permission to copy this thesis for scholarly purposes may be granted by the
Director of the Library or my major professor. It is understood that any copying or
publication of this thesis for financial gain shall not be allowed without my further
written permission and that any user may be liable for copyright infringement.
Agree (Permission is granted.)
Kevin Moore__________________________________ ________________ Student Signature Date Disagree (Permission is not granted.) _______________________________________________ _________________ Student Signature Date
53