Content uploaded by Norbert Radacsi
Author content
All content in this area was uploaded by Norbert Radacsi on Jul 05, 2018
Content may be subject to copyright.
Research Article
Electrospray Crystallization for High-Quality
Submicron-Sized Crystals
Nano- and submicron-sized crystals are too small to contain inclusions and are,
therefore, expected to have a higher internal quality compared to conventionally
sized particles (several tens to hundreds of microns). Using electrospray crystalli-
zation, nano- and submicron-sized crystals can be easily produced. With the aid
of electrospray crystallization, a mist of ultrafine solution droplets is generated
and subsequent solvent evaporation leads to crystallization of submicron-sized
crystals. Using cyclotrimethylene trinitramine (RDX) solutions in acetone, the
conditions for a stable and continuous jet were established. At relatively small
nozzle diameters and relatively low potential differences, hollow spheres of RDX
crystals were observed. At a higher nozzle diameter and potential difference and
in the region of a continuous jet, RDX crystals with an average size of around
400 nm could be produced. In order to test the quality of the submicron-sized
energetic material, impact and friction sensitivity tests were carried out. The test
results indicate that the submicron-sized product had reduced friction sensitivity,
indicating a higher internal quality of the crystalline product.
Keywords: Crystallization, Cyclotrimethylene trinitramine, Electrospray crystallization,
Nanoparticles
Received: December 09, 2010; revised: January 20, 2011; accepted: January 21, 2011
DOI: 10.1002/ceat.201000538
1 Introduction
The product quality of crystalline materials is determined by
the crystallization process applied to produce these materials.
The crystalline product quality is in turn determined by the
crystal size distribution, kind of solid state, morphology, pur-
ity, and internal defects. There are several methods to optimize
the crystallization conditions in conventional crystallization
processes in order to arrive at a certain quality of the crystal-
line product. Using a specific geometry with low cooling rate
can end up in producing 1-mm crystals with narrow crystal
size distribution [1].
Inclusions, dislocations, and point, line, or surface defects
are internal defects [1]. For energetic materials it is thought
that the initiation of an explosion can be caused by an unin-
tentional shock wave hitting a defect upon its propagation
through a crystal. It has been demonstrated that the higher the
product quality of a solid energetic material, the less sensitive a
plastic-bonded explosive containing this energetic material
becomes [2]. A reduced sensitivity means that the detonation
threshold is higher which would prevent unintentional detona-
tions, therefore, these explosives are safer to handle. For the
energetic material cyclotrimethylene trinitramine (RDX), for
instance, a reduced cooling rate in batch cooling crystalliza-
tions resulted in RDX crystals with decreased inclusion con-
tent [3]. Therefore, optimizing conventional crystallization
conditions is a good approach to achieve lower internal defect
densities and thus less sensitive energetic materials.
Another approach is to greatly reduce the crystal size. Since
inclusions are relatively large, submicron-sized crystals are
believed to contain a smaller amount of inclusions. Crystal
defects like dislocations usually originate from inclusions and,
therefore, a much lower overall defect content is expected in
nano- and submicron-sized crystals. Due to the relation of
sensitivity and defect content of energetic compounds, the
product quality can be easily quantified with the help of sensi-
tivity tests.
In this study, an electrospray crystallization method for pro-
ducing submicron-sized crystals of the energetic compound
RDX is presented. The method combines electrospraying of an
undersaturated solution and solvent evaporation from the cre-
ated droplets. The extent to which submicron particles can be
produced with electrospray crystallization as well as their sen-
sitivity performance as energetic material were investigated.
www.cet-journal.com © 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Eng. Technol. 2011,34, No. 4, 624–630
Norbert Radacsi
1
Andrzej I. Stankiewicz
1
Yves L. M. Creyghton
2
Antoine E. D. M. van der
Heijden
1,2
Joop H. ter Horst
1
1
Delft University of Technology,
Process and Energy
Laboratory, Intensified
Reaction and Separation
Systems, Delft,
The Netherlands.
2
TNO Defence, Security and
Safety, Rijswijk, The
Netherlands.
–
Correspondence: N. Radacsi (n.radacsi@tudelft.nl), Delft University of
Technology, Process and Energy Laboratory, Intensified Reaction and
Separation Systems, Leeghwaterstraat 44, NL-2628 CA Delft, The
Netherlands.
624
2 Electrospray Crystallization
There are several techniques for producing nanoparticles. Con-
ventional methods like milling or pyrolysis and advanced
methods like using thermal plasma [4], rapid expansion of
supercritical solutions [5], or spray drying [6] are often used
for nanoparticle production. Electrospray crystallization is one
of the advanced methods to achieve nano- and submicron-
sized crystals [7]. A constant potential difference is applied
between a grounded plate and a metal capillary (the nozzle)
through which a conductive solution is pumped. If the poten-
tial difference is sufficiently high, electrostatic forces overcome
the surface tension and any flow arriving at the tip of the noz-
zle is directly emitted as a jet of liquid droplets from a Taylor-
cone formed at the nozzle [8]. Due to the potential difference
and the use of a conducting solution the surface of the droplets
is charged. The surface charge density is mainly determined by
the potential difference.
The jet contains small droplets which are accelerated to-
wards the grounded plate by the electric field. Upon using a
sufficiently volatile solvent such as acetone, solvent evapora-
tion occurs which increases the surface charge because of the
decrease in droplet volume. As the surface charge reaches a
critical value (Rayleigh limit [9]), electrostatic forces overcome
the surface tension, and the droplet disrupts into smaller drop-
lets to reduce the surface charge density by creating more sur-
face area. This disruption process is called Coulomb fission
[10].
The solute concentration increases as the solvent evaporates
from the droplet. At some point during this process of droplet
evaporation and disruption, the driving force for crystalliza-
tion becomes sufficiently large for crystal nucleation and
growth to occur. It is assumed that crystallization is confined
to the volume of the droplet [11]. Therefore, if the droplets are
sufficiently small, typically one crystal per droplet is formed.
Aggregation of droplets is prevented because the droplets have
surface charges of equal sign. Thus, submicron or even nano-
sized crystals can be formed. These submicron crystals accu-
mulate at the grounded surface loosing their surface charge.
Fig. 1 shows the scheme of the electrospray crystallization pro-
cess and the investigated process parameters (nozzle diameter
d, flow rate u, potential difference DU, solution concentration
c, and working distance D).
3 Experimental
3.1 Materials
Class 2 RDX with an average size of around 400 lm was pur-
chased from Chemring Nobel A.S., Norway. For the electro-
spray crystallization different solution concentrations were
prepared with 99.8 % acetone obtained from Merck. Care was
taken to choose materials in the device that withstand the
exposure to acetone.
3.2 Solubility and Crystal Growth Measurements
The Crystal16 developed by Avantium Technologies BV was
used to determine the RDX solubility in acetone. With Crys-
tal16, cloud points and clear points of sixteen 1-mL solution
aliquots can be measured in parallel and automatically, based
on turbidity. The temperature at the point where the suspen-
sion becomes a clear solution (clear point) upon slow heating
(0.3 °C min
–1
) was taken as the saturation temperature of the
measured sample, of which the composition was established
beforehand. The cloud point refers to the point at which solid
material first appears upon a decrease of temperature.
The crystal growth rate was studied in a stirred cooling crys-
tallization process. For this, the Crystalline Particle Viewer,
also developed by Avantium Technologies BV, was used to
visualize the suspension during cooling crystallization of RDX
with an on-board camera. This multiple 3-mL reactor setup
combines turbidity measurements with four independent real-
time particle visualization modules. This device is capable to
take images every second, and its software analyzes and calcu-
lates particle size, distribution, and shape. The crystalline soft-
ware calculates the median crystal size for every image which
was taken every second. Each minute an average crystal size
was determined from these images. The average crystal growth
rate during the crystallization process can then be
estimated by dividing the size increase and the
time needed to obtain that size increase. Suspen-
sions (3 mL) were prepared of RDX in acetone.
The samples were heated above their saturation
temperature in order to obtain a clear solution.
Then they were cooled down to –15 °C using a
cooling rate of 0.5 °C min
–1
.
3.3 Single-Nozzle Electrospray
Crystallization Setup
The electrospray crystallization setup (see Fig. 2)
consisted of an Aitecs SEP-10S Syringe Pump with
a 50-mL plastic syringe. A Wallis ±10 kV DC power
supply was used to provide the potential difference
DUbetween the tip of the nozzle and the grounded
Chem. Eng. Technol. 2011,34, No. 4, 624–630 © 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.cet-journal.com
Figure 1. Process scheme of electrospray crystallization. Varied operation condi-
tions: nozzle diameter (d), flow rate (f), concentration (c), potential difference
(DU), and working distance (D). Solvent evaporation, Coulomb fission, and crys-
tallization occur during the process.
Electrospray crystallization 625
plate which were separated at a working distance D. The pump
and a Festo 6x1 pneumatic tube were used to transport the
solution with concentration cto the nozzle with a certain flow
rate u. Seven different nozzles (EFD, USA) were used for the
experiments. All of them had the same length (25.4 mm) but
varying inner diameters d(0.1524, 0.254, 0.3302, 0.4064,
0.508, 0.5842, or 1.3716 mm, referred to in the paper as, re-
spectively, the 0.15, 0.25, 0.33, 0.4, 0.51, 0.58mm nozzle for
convenience). The experiments were performed at room tem-
perature.
3.4 Multiple-Nozzle Electrospray Crystallization
Setup
Multiple nozzles have the advantage that higher production
rates can be achieved. An eight-nozzle system was used with a
Meredos TL-EAD peristaltic pump in order to have an
equal distribution of the solution flow over the nozzles
(Fig. 2). The rest of the setup was similar to the single noz-
zle setup. The experiments were performed at room tem-
perature.
3.5 Characterization Tests
For morphological and size characterization of the crystal-
line samples a Philips XL30 FEG scanning electron micro-
scope was applied. Typical instrument settings of the elec-
tron beam were 2 kV, spot size 3, and working distance
10 mm. Samples were investigated as is, without any
further treatment. Only low electron beam energies were
used since energies higher than 2 kV at higher magnifica-
tions resulted in decomposition of the energetic materials.
The BAM Fallhammer and Friction Apparatus tests are
the recommended test methods in the UN recommenda-
tions for the transport of dangerous goods [12]. For the
impact sensitivity test a BAM Fallhammer device was used
(according to the international guidelines EC A.14 and UN
3(a) (ii)) [12]. The temperature during the test was 23 °C.
The volume of the submicron-RDX sample was approxi-
mately 40 mm
3
and the drop weight was 5 kg. During the
test, the drop height was decreased from 50 cm until initia-
tion of the sample did not occur for six times in series.
The sensitivity to friction was determined by a BAM
Friction Apparatus (according to the international guide-
lines EC A.14 and UN 3(b) (i) [12]. The sample is placed
on a roughened porcelain plate which is rigidly attached to
the sliding carriage of the friction apparatus. A cylindrical
porcelain peg is placed on top of the sample. The porcelain
plate can move forward and backward under the porcelain
peg. The smallest load of the peg, which causes deflagra-
tion, crackling, or explosion of the test sample at least once
in six consecutive trials, is the outcome of the friction sen-
sitivity test. The quantity of the test sample is 10 mm
3
.
4 Results and Discussion
4.1 Solubility and Conventional Growth Rate
Fig. 3 shows the solubility line of RDX in acetone determined
from the saturation temperature measurements carried out
with Crystal16. At 45 °C the solubility for RDX in acetone is
111 mg mL
–1
. The RDX solubility in acetone increases with
temperature.
Cooling crystallization experiments were performed with
the Crystalline Particle Viewer setup. During cooling crystalli-
zation, the solution was monitored in situ by measuring the
transmission of light through the solution and by taking im-
ages simultaneously. The evolving crystal size was monitored
after a sufficient amount of crystals was present. Since the res-
olution of the on-board camera is sufficiently high (5–10 lm/
pixel) to identify crystals above 30 lm size, only these were
taken into account. Fig. 4 presents the average size of the RDX
crystals as a function of time. Initially the size seems to be con-
www.cet-journal.com © 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Eng. Technol. 2011,34, No. 4, 624–630
Figure 2. Single-nozzle electrospray crystallization setup on the left and
the eight-nozzle experimental setup on the right showing the capillary
nozzle (A), high-voltage connectors (B), grounded plate of aluminum
foil on which the particles are collected (C), the nozzle holders (D), and
the Meredos TL-EAD peristaltic pump (E).
Figure 3. Solubility of RDX in acetone as a function of temperature T.
626 N. Radacsi et al.
stant slig htly above 30 lm. After around 10 min, the size seems
to increase roughly linearly with time, leading to an average
growth rate of about 0.061 lms
–1
for RDX in acetone (see
Fig. 4). These values are typical for the crystal growth of or-
ganic compounds from solutions
4.2 Single-Nozzle Electrospray Crystallization
Operation Conditions
For RDX, acetone is a favorable solvent to use since RDX crys-
tals from acetone have a compact shape [13]. Moreover, ace-
tone is a (slightly) conductive solvent which allows to electro-
spray the solution. Furthermore, acetone has a high vapor
pressure, resulting in a relatively high evaporation rate during
electrospray crystallization.
The first goal was to find the process conditions leading to
the formation of a continuous jet in the single nozzle setup. A
continuous jet is established when there is a stationary droplet
emission from the nozzle without any interruption. Then, the
effect of the five process variables was studied: the flow rate u,
potential difference DU, working distance D, nozzle diameter
d, and the solution concentration c.
Using a solution with a concentration of 20.8mg mL
–1
RDX,
a continuous jet was observed at a potential difference of
+4.8 kV, a nozzle diameter of 0.58mm, a working distance of
25 mm, and a flow rate of 2.8 mL h
–1
. This is taken as our refer-
ence experiment for studying the operation conditions as well
as the crystalline product characteristics.
4.2.1 Flow Rate
A continuous jet was observed at flow rates between 1 and
5mLh
–1
and otherwise equal settings compared to the stan-
dard experiment. At flow rates below 1 mL h
–1
, the solution
was only intermittently sprayed from the nozzle. Apparently
the flow rate then is too small to achieve a steady-state electro-
spray of droplets from the tip. Flow rates higher than 5 mL h
–1
resulted in formation of an unwanted continuous flow of solu-
tion from the tip without jet formation: The electrospraying
cannot keep up with the accumulation of solution at the tip
due to the relatively high flow rate.
4.2.2 Potential Difference
A continuous jet is obtained when the potential difference is
varied between 3.8 and 4.9 kV while using the standard settings
for the other variables. Applying a potential difference above
4.9 kV resulted in formation of multiple jets. Three jets were
observed when using a potential difference of 6.2 kV with an
RDX solution of 20.8 mg mL
–1
, a working distance of 25 mm,
and a flow rate of 2.8 mL h
–1
. Multiple jets are caused by the
formation of multiple Taylor cones. These multiple Taylor
cones form because of the extremely high charge density at the
nozzle tip.
Fig. 5 shows the minimum and maximum potential differ-
ence for obtaining a continuous jet for different working dis-
tances using an RDX solution concentration of 20.8 mg mL
–1
,
a flow rate of 2.8 mL h
–1
, and a nozzle diameter of 0.58 mm.
The area between the two lines indicates the region for obtain-
ing a continuous jet.
4.2.3 Working Distance
Fig. 5 also shows that at larger working distances a higher
potential difference is needed to obtain a continuous jet. The
working distance between the nozzle tip and the grounded
plate was varied between 5 and 70 mm. Working distances
smaller than 5 mm were not tested since unwanted sparks
might occur. At working distances smaller than 10 mm solvent
evaporation of the droplets travelling from the tip to the
grounded plate was only partial, and the remaining acetone
evaporated on the grounded plate. This is unwanted since
crystallization then partly occurs at the grounded plate. At
working distances above 35 mm, the developed jet was contin-
Chem. Eng. Technol. 2011,34, No. 4, 624–630 © 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.cet-journal.com
Figure 4. Left: RDX crystals during cooling crystallization from acetone. Right: median crystal size (D
50
) as a function of time. The cooling
rate during the experiment was set to 0.5 °Cmin
–1
.
Electrospray crystallization 627
uous but the droplets were not homogeneously distributed in
time and space within the jet, i.e., it was observed that after a
time interval of several seconds a part of the flow was emitted
in relatively large droplets from the nozzle tip which travelled
at the outer side of the jet to the ground plate. These droplets
often partly evaporated on the collector plate.
4.2.4 Nozzle diameter
The minimum and maximum potential difference for the for-
mation of a single continuous jet were determined for seven
different nozzle diameters (Fig. 6). For the concentration and
flow rate the standard values were used, while 35 mm was cho-
sen for the working distance since that was the upper limit
below which no dripping occurred. A larger nozzle diameter
demanded a higher potential difference in order to obtain a
continuous jet.
4.2.5 Concentration
While using the standard settings for the other variables, de-
creasing the RDX concentration did not affect the continuous
jet formation: For concentrations between 2 and 20.8 mg mL
–1
a continuous jet was formed. However, increasing the concen-
tration led to unwanted crystallization at the nozzle tip. This
eventually led to blocking of the nozzle. Using a concentration
of 24.9 mg mL
–1
, crystals were growing on the nozzle tip
which blocked the flow within 15 min. Using a solution
close to the supersaturated concentration
(41.6 mg mL
–1
RDX) caused the blocking
to occur within 40 s.
4.3 Product Characterization
For checking the product size and shape,
scanning electron microscopy was used.
The resulting submicron-sized crystals of
the standard experiment are shown in
Fig. 7. The produced crystals have a spher-
ical shape and a size ranging from 200 to
600 nm. No extensive agglomeration was
observed which indicates that under these
conditions droplets are not aggregating
and the small droplet volume causes the
nucleation of only a single RDX crystal per
droplet. This single crystal per droplet then
consumes the dissolved RDX by growth
upon further evaporation of the acetone in
the droplet. A crystal size of 400 nm indi-
cates that the solution droplet from which
the crystal is formed is around 1.77 lm.
These experiments show that electrospray
crystallization can be used to obtain sub-
micron-sized crystals of RDX.
4.3.1 Hollow-Sphere Agglomerates
When the potential difference was decreased to below 4.5 kV
or lower compared to the standard experiment, a considerable
amount of hollow spheres of multiple agglomerated RDX crys-
tals were observed in the crystalline material (Fig. 8). These
hollow spheres probably develop due to the combination of a
www.cet-journal.com © 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Eng. Technol. 2011,34, No. 4, 624–630
Figure 5. Potential difference DUand working distance Dto obtain a stable and continu-
ous single jet using a c= 20.8 mg mL
–1
RDX solution in acetone, a nozzle diameter of
d= 0.4 mm, and a flow rate of u= 2.8 mL h
–1
. Upper line: potential difference above which
multiple jets were obtained; lower line: potential difference below which intermittently
sprays were emitted from the nozzle tip. Area between the lines: region for continuous
jet formation. Black point: conditions of the reference experiment.
Figure 6. Minimum and maximum potential differences DUfor
obtaining a single continuous jet as a function of the nozzle
diameter dat a working distance D= 35 mm, concentration
c= 20.8 mg mL
–1
, and flow rate u= 2.8 mL h
–1
. In the region be-
tween the lines a continuous jet was formed.
628 N. Radacsi et al.
relatively low potential difference and a relatively high RDX
concentration. Under these conditions, relatively large droplets
might form due to a decreased surface charge, less Coulomb
fission, and more aggregation of droplets. Due to the large
concentration, these droplets become supersaturated relatively
soon after being sprayed from the nozzle. Then multiple crys-
tals can form at the surface of the droplet which develop in a
shell of connected elongated crystals of which the internal is
empty. An alternative explanation for forming similar particles
can be found elsewhere [14].
At an RDX concentration c= 20.8 mg mL
–1
, these hollow-
sphere agglomerates could be prevented by using a combina-
tion of a high potential difference (DU≥+4.5 kV) and a large
nozzle diameter (d≥0.58 mm). These boundar y values for
nozzle diameter and potential difference decrease when lower
concentrations are used. This indicates that also decreasing the
solution concentration can prevent the presence of these hol-
low-sphere agglomerates in the product.
In case the potential difference is higher than +4.5 kV, the
hollow spheres were not observed, also above the potential dif-
ference threshold for a continuous single jet. At higher poten-
tial differences the charge density on the droplet surface is larg-
er which might result in smaller droplet sizes after Coulomb
fission. In smaller droplets the chance of forming only one
crystal is higher. Moreover, at higher surface charges the aggre-
gation of droplets is decreased.
With 20.8 mg mL
–1
concentration, hollow spheres and ag-
glomerates can be observed in case the potential difference is
rather low. In Fig. 5 it can be seen that the operation window
for potential difference is around 1 kV for obtaining a continu-
ous jet. According to our measurements with this relatively
high concentration, hollow spheres can be observed in the low-
er region of this window.
The formation of these hollow-sphere agglomerates makes
the operation window for submicron-sized RDX production
without agglomerates produced from a continuous jet rather
small. To obtain submicron-RDX crystals from a continuous
jet, a low concentration (c≤20.8 mg mL
–1
) and a high potential
difference (DU≥+4.5 kV) are recommended with a nozzle
diameter of d≥0.58 mm at a working distance of D≤35 mm
and a flow rate of 2.8 mL h
–1
.
4.3.2 Crystal Growth Rate during Electrospray Crystallization
The average growth rate in the electrospray crystallization pro-
cess was calculated by dividing the product crystal size
(400 nm) with the estimated time the solvent needed for evap-
oration. From the experiments with different working dis-
tances it was observed that working distances below 10 mm
resulted in acetone evaporation partly on the collector plate.
Therefore, 10 mm was chosen as the working distance for cal-
culating the time for solvent evaporation. The actual time of
growth of the nanoparticle is smaller since the nanoparticle is
formed somewhere along the path from nozzle to plate. In
order to calculate the time, the droplet velocity is also needed.
At a voltage of 5 kV in an electrospraying process the droplet
velocity of acetone was reported to be 10.8 m s
–1
[15]. The time
the droplet is travelling was thus estimated to be 1.08 ms.
Thus, the estimated growth rate was 370 lms
–1
. Compared
with cooling crystallization, electrospray crystallization results
in at least 6000 times higher growth rates which seems very
large.
4.4 Multiple-Nozzle Electrospray Crystallization:
Product Sensitivity
To check the expected product quality improvement, impact
and friction sensitivity tests were performed. For these tests, at
least 1 g of material was needed and, therefore, the multiple
nozzle setup was used to increase the production rate. The
production rate with the eig ht-nozzle setup was 84 mg h
–1
using 24.9 mg mL
–1
RDX at a flow rate of 1.8 mL h
–1
.
Chem. Eng. Technol. 2011,34, No. 4, 624–630 © 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.cet-journal.com
Figure 7. 200–600 nm sized RDX crystals from the standard elec-
trospray crystallization experiment using the 0.58-mm nozzle at
a potential difference of +4.8 kV, a working distance of 25 mm, a
solution concentration of 20.8 mg mL
–1
, and a flow rate of
2.8 mL h
–1
.
Figure 8. Hollow spheres of RDX crystals produced with the
0.58-mm nozzle at a potential difference of 4.2 kV, 25 mm work-
ing distance, a solution concentration of 20.8 mg mL
–1
, and a
flow rate of 2.8 mL h
–1
. Typically the size of these spheres was
around 4 lm.
Electrospray crystallization 629
With the produced samples, impact sensitivity tests were
carried out. When the drop height was set at 15 cm w ith a 5-kg
drop weight (7.5 J impact energy), there was no ignition using
the produced submicron-sized RDX. Ignition only occurred
when the height was set to 20 cm (10 J). The conventional
RDX needs 7.5 J energy for ignition [12] which is not signifi-
cantly different from that measured for the submicron-sized
RDX.
Friction sensitivity tests were performed as well with nano-
crystals of RDX. Tests were carried out first at the highest
possible load (360 N) to check the friction sensitivit y of the
400-nm RDX. For this energetic material even with this high
load no ignition occurred. Conventional RDX needs 120 N
load for ignition [12]. Apparently, the friction sensitivity is
much lower for the submicron-sized energetic particles than
for the conventionally sized explosive crystals. These sensitivity
tests might indicate that the submicron-sized crystals contain
less defects and, therefore, have a higher internal quality.
5 Conclusions
Submicron-sized crystals of RDX can be produced using elec-
trospray crystallization. The sensitivity of samples having an
average size of around 400 nm for RDX was tested. The sam-
ples were remarkably insensitive to friction stimuli while an
insignificant difference for the impact sensitivity was observed.
The operation window to establish a continuous jet and
produce submicron-sized crystals is relatively narrow, but
experimentally feasible. It is being worked on to optimize this
process to higher yields and improved process control using a
new process design.
Acknowledgment
We would like to thank Leon Wassink for performing some ex-
periments, Willem Duvalois for the SEM analysis, and Ilse
Tuinman for providing the peristaltic pump.
The authors have declared no conflict of interest.
References
[1] J. W. Mullin, Crystallisation, 4th ed., Butterworth-Heine-
mann, Woburn, MA 2001.
[2] A. E. D. M. van der Heijden, Y. L. M. Creyghton, E. Marino,
R. H. B. Bouma, J. H. G. Scholtes, W. Duvalois, Propellants,
Explos., Pyrotech. 2008,33 (1), 25.
[3] J.-W. Kim, J.-K. Kim, H.-S. Kim, K.-K. Koo, Cryst. Growth
Des. 2009,9 (6), 2700.
[4] N. Rao et al., Plasma Chem. Plasma Process. 1995,15 (4),
581.
[5] V. Stepanov, L. N. Krasnoperov, I. B. Elkina, X. Zhang, Pro-
pellants, Explos., Pyrotech. 2005,30 (3), 178.
[6] H. Qiu, V. Stepanov, A. R. Di Stasio, T. Chou, W. Y. Lee,
J. Hazard. Mater. 2011,185, 489.
[7] A. Jaworek, Powder Technol. 2007,176, 18.
[8] G. I. Taylor, Proc. R. Soc. London, Ser. A 1964,280, 383.
[9] L. Rayleigh, Phil. Mag. 1882,14, 184.
[10] K. Suzuki, H. Matsumoto, M. Minagawa, M. Kimura, A. Ta-
nioka, Polym. J. 2007,39 (11), 1128.
[11] E. Revalor, Z. Hammadi, J. P. Astier, R. Grossier, E. Garcia,
C. Hoff, K. Furuta, T. Okutsu, R. Morin, S. Veesler, J. Cryst.
Growth 2010,312, 939.
[12] R. Meyer, J. Köhler, A. Homburg, Explosives, 6th ed., Wiley-
VCH, Weinheim 2007.
[13] J. H. ter Horst, R. M. Geertman, A. E. D. M. van der Heijden,
G. M. van Rosmalen, J. Cryst. Growth 1999,198/199, 773.
[14] Y. Jiang, J. Electrochem. Soc. 2007,154 (8), E107.
[15] H. Oh, K. Kim, S. Kim, Aerosol Sci. 2008,39, 801.
www.cet-journal.com © 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Chem. Eng. Technol. 2011,34, No. 4, 624–630
Table 1. Impact and friction sensitivity of 400-nm RDX compared
with conventionally sized RDX.
Impact sensitivity [J] Friction sensitivity [N]
Conventional RDX 7.5 120
Submicron-sized RDX 10 > 360
630 N. Radacsi et al.
