Examining explosives handling sensitivity of

trinitrotoluene (TNT) with different particle

sizes
Cite as: AIP Conference Proceedings 2272, 050015 (2020); https://doi.org/10.1063/12.0001125
Published Online: 04 November 2020

Virginia W. Manner, Clay G. Tiemann, John D. Yeager, et al.

ARTICLES YOU MAY BE INTERESTED IN

Small-scale explosives sensitivity saftey testing: A departure from Bruceton

AIP Conference Proceedings 1426, 713 (2012); https://doi.org/10.1063/1.3686378

Understanding the shock and detonation response of high explosives at the continuum and
meso scales
Applied Physics Reviews 5, 011303 (2018); https://doi.org/10.1063/1.5005997

Synthesis and sensitivity studies of ETN and PETN derivatives

AIP Conference Proceedings 2272, 050013 (2020); https://doi.org/10.1063/12.0001121

AIP Conference Proceedings 2272, 050015 (2020); https://doi.org/10.1063/12.0001125 2272, 050015

© 2020 Author(s).
 Examining Explosives Handling Sensitivity of
Trinitrotoluene (TNT) with Different Particle Sizes
Virginia W. Manner1,a) Clay G. Tiemann,1 John D. Yeager,1 Lisa M. Kay,1
Nicholas Lease,1 Marc J. Cawkwell,1 Geoff W. Brown,1 Spencer P. Anthony,1
and Danielle Montanari1
1
Los Alamos National Laboratory, Los Alamos, NM 87545, USA

a)
Corresponding author: [email protected]

Abstract. It is critical to understand and mitigate for explosive sensitivity, from applications related to stockpile
maintenance and development, to law enforcement safety, to basic research. The handling sensitivity of explosives is
controlled by factors that span from fundamental intramolecular effects to mesoscale structure and defects. Although the drop-
weight impact test has been used to characterize energetic material sensitivity for over 50 years, there is still very limited
understanding on which features of explosives influence or control impact sensitivity. Fundamental chemistry is likely the most
important player in handling sensitivity of energetics, however secondary effects such as particle size and morphology are
thought to influence the results in impact sensitivity tests. We have recently discovered that batches of trinitrotoluene (TNT)
obtained with a freshly synthesized method versus a historical lot sample exhibit significantly different impact sensitivity
values. To determine if differences in particle size were the reason for the different results, in the present work we have prepared
13 different samples of TNT and tested each for particle size and impact sensitivity. Though the particle sizes of the samples
range from 44 – 1502 µm, impact test values fall mostly within error, given the large standard deviation in the test. The historical
lot TNT sample is an exception, and is more sensitive than the newly prepared batches. Overall, more work is needed to
determine if other factors such as purity and particle morphology (rather than size) are more important for influencing impact test
results in TNT.

INTRODUCTION

Understanding and tuning explosives sensitivity is an important area of concern, as explosive accidents frequently
occur around the world. Drop-weight impact tests, which drop a 2.5 kg weight onto an explosive sample, are
generally used to rank explosives in terms of sensitivity to non-shock ignition. We have been investigating different
categories of explosives to determine what factors influence handling sensitivity, and have discovered three
important regimes to consider. Basic chemistry of energetic functional groups appears to have the most
important influence on sensitivity, followed by intermolecular interactions such as crystal packing characteristics
and hydrogen bonding. Finally, there have been a significant number of studies investigating the importance of
particle size effects on handling sensitivity. We have recently studied a system of derivatives of pentaerythritol
tetranitrate (PETN), where quantum molecular dynamics modeling indicates that initial reactions are very similar
among these explosives with the same nitrate ester energetic functional group. Though all of the nitrate ester
explosives in the study exhibited similar sensitivity to PETN, we could make small adjustments in impact test
values by modifying properties such as crystal packing, compressibility, and hydrogen bonding [1,2]. In order to
consider explosives systems with different functional groups, we have recently investigated a trinitrobenzene (TNB)
system [3]. Surprisingly, in that study, two different samples of trinitrotoluene (TNT) exhibited significantly
different sensitivities from one another (one was a historical lot sample of TNT, and one was a sample recently
synthesized in our labs). To examine if particle size influences impact test results with TNT, in the present work
we have prepared 13 different TNT samples and tested each for impact sensitivity and particle size.

Shock Compression of Condensed Matter - 2019

AIP Conf. Proc. 2272, 050015-1–050015-6; https://doi.org/10.1063/12.0001125
Published by AIP Publishing. 978-0-7354-4000-5/$30.00

050015-1
 RESULTS AND DISCUSSION
All TNT samples were prepared starting from a LANL-obtained historical lot batch of TNT. This TNT was either
recrystallized using different solvents and methods, or heated in the presence of water to melt-cast the material.
Some samples were sieved to further control the size distribution. Table I shows a summary of preparation
conditions for all samples, and detailed information is given in the Experimental Section. After the samples were
prepared, they were tested to determine drop-weight impact sensitivity, run through a laser diffraction particle
size analyzer, and imaged using scanning electron microscopy. Table I shows the mean particle size, with the size
distribution within 1s for each sample. A wide range of particle sizes was obtained for the recrystallized and
melt-cast samples, with most exhibiting large standard deviations. Figure 1 shows the 50% drop height values
(DH50) for each sample, plotted against the particle size. All of the recrystallized and melt-cast samples exhibited
DH50 values ranging between 204 cm and 319 cm. However, the original lot TNT sample has a significantly lower
drop weight impact height than the other samples, with a DH50 of 121 ± 25 cm.

TABLE I. Description of TNT recrystallization methods, along with particle size data and drop-weight impact data.

Sample Number Method Particle Size (µm) 50% Drop

Height (cm)
1 Slow Recrystallization MeOH 861 ± 474 216 ± 22
2 Fast Recrystallization EtOH 402 ± 382 204 ± 30
3 Crash MeOH/Water 49 ± 81 210 ± 27
4 Slow Acetone/Ice 180 ± 177 251 ± 14
5 Slow Acetone/Ice 194 ± 168 220 ± 19
6 Slow Acetone/Ice 93 ± 109 214 ± 15
7 Melt-Cast Prill 488 ± 194 228 ± 20
8 Melt-Cast Prill 77 ± 45 229 ± 18
9 Melt-Cast Prill 44 ± 26 319 ± 6
10 Lot 1366 ± 408 121 ± 25
11 Melt-Cast Prill 1502 ± 455 208 ± 25
12 Melt-Cast Prill 1485 ± 352 238 ± 53
13 Melt-Cast Prill 1076 ± 377 277 ± 7
14 Melt-Cast Prill 118 ± 71 244 ± 40

SEM images were taken for all samples to compare particle morphology and broadly verify the particle size
measurements. Figure 2 shows typical images for each preparation condition. Generally, the smaller crystals are
prismatic or elongated prisms, although melt-cast particles appear slightly rounder or rougher. In most of the samples
with large particle size, the largest particles are somewhat rounded and rough. However, the lot TNT (Sample # 10,
DH50 = 121 ± 25 cm) shows two different size regimes. Figure 2 clearly shows that the large “particles” are actually
agglomerates with surface features that closely resemble the small particles of Sample # 3 (DH50 = 210 ± 27 cm) and
Sample # 6 (DH50 = 214 ± 15 cm). By contrast, the large particles of Sample # 12 (DH50 = 238 ± 53 cm) are clearly
solid particles with small amounts of other TNT crystal debris coating the surface. Overall, with the exception of the
Lot TNT sample, variations in preparation and control of size distribution through sieving and recrystallization
conditions do not appear to greatly influence impact sensitivity of the TNT samples.
In other explosives systems, there is evidence that particle size and shape can influence both shock and impact
sensitivity. For example, recent studies have shown that nano-RDX (hexahydro-1,3,5-trinitro-1,3,5-triazine) and
HMX (octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine) demonstrate lower sensitivity than the larger micron-scale
explosives [4], and RDX powders with different particle sizes and distributions correlate strongly with impact
sensitivity [5]. The size of lead azide crystals has been shown to correlate with impact sensitivity as well [6].
However, ultrafine PETN has been shown to be less sensitive to ignition in impact tests than conventional material
[7], and we have observed that adjusting particle size of erythritol tetranitrate (ETN) does not appear to influence drop
weight impact test results [8]. Both ETN and TNT melt at low temperatures, so it is possible that the low melting point
of TNT could cause melting on the timescale of the impact test, causing particle size to be less important. We have
recently determined that melting point may influence impact sensitivity in a system of nitrate esters with modified
melt temperatures [9].

050015-2
 The agglomerate nature of Sample # 10 (Lot TNT) could explain the higher impact sensitivity when compared to
the other samples. As mentioned in previous work [2,10], there are many possible energy transfer mechanisms at play
in the drop-weight impact test. Depending on the strength of the interaction between the many small TNT crystals,
the large agglomerate may effectively behave as a highly defect-ridden “single crystal”, where the extensive porosity
serves as nucleation sites for localized heating and non-uniform deformation. This hypothesis will be tested in future
work by grinding or otherwise breaking apart the agglomerate Lot TNT and repeating the impact tests. Another
possibility is that the Lot TNT has some residual impurity which causes increased sensitivity; such an impurity, if it
exists, would likely be removed during the extensive recrystallization processes used on all of the other samples. This
possibility will also be examined in future efforts.

FIGURE 1. Drop-weight impact data for TNT samples versus the average particle size determined from Coulter particle size
measurements. The best fit line is drawn through the melt-cast prilled samples (blue circles) and recrystallized samples (grey
squares). The lot TNT (blue diamond) is not included in the best-fit line.

FIGURE 2. Representative images from Samples 3 (crash precipitate), 6 (slow recrystallization), 10 (lot), and 12 (melt-cast prills).

050015-3
 EXPERIMENTAL
TNT Crystallization Procedures. 2,4,6-Trinitrotoluene (TNT) is toxic and classified as 1.1D high explosive. The
following experiments should only be conducted in a laboratory qualified/equipped to work with energetic materials
and performed by trained personnel. The TNT used in this experiment was acquired from LOT BC-455141/4. All
reagents used were ACS reagent grade or better and were used as received.

Stock TNT for Crystallization. To a 500 mL Erlenmeyer flask equipped with a magnetic stir bar was added
approximately 50 g of lot TNT and enough acetone to dissolve it at room temperature. To the solution was added a
few spatula tips of powdered activated carbon. The mixture was covered and stirred for 2 hr. The mixture was then
vacuum filtered over a fritted funnel loaded with Celite. The filtrated was then dumped into 1 L of deionized water to
precipitate the TNT. The TNT was collected by vacuum filtration, washed with water, and placed in a vacuum
desiccator to dry over the course of several days. This TNT was then used for further crystallization experiments.

Recrystallization from Hot Solvent. Method 1: 5 g of stock TNT was recrystallized from hot methanol. The solution
(in a Parafilm sealed beaker) was allowed to cool very slowly in an insulated container overnight. This resulted in
formation of relatively long (~1 cm) clear very light tan (almost colorless) needles. The material was vacuum filtered
and placed in a vacuum desiccator to dry. The TNT was used as prepared.
Method 2. 5 g of stock TNT was recrystallized from hot ethanol. The solution cooled quickly in an ice bath with
vigorous stirring. This resulted in formation of small clear very light tan (almost colorless) needles. The material was
vacuum filtered and placed in a vacuum desiccator to dry. The TNT was used as prepared.

Crash Precipitation. 5 g of stock TNT was dissolved in hot methanol. A 1 L beaker was charged with 700 mL of
deionized water and equipped with an overhead stirrer. The water was cooled to 8 °C and stirring speed was set to 450
rpm. The hot TNT solution was then poured (keeping the stream as small as possible but continuous) into the vortex.
This resulted in a very flocculent white powder. The material was vacuum filtered and placed in a vacuum desiccator
to dry. The TNT was used as prepared [11].

Solvent/Anti-solvent Recrystallization. To a 1 L beaker was added 15 g of stock TNT, 200 mL of acetone, and a
large stir bar. The beaker was placed on a hot place set to 55 °C (this was to maintain a solution temperature of 20 -
25 °C) and stirring was set to the lowest setting (~60 rpm). A single ice cube was placed in the TNT/acetone solution
and allowed to completely melt. This process was repeated and eventually crystals started to form. By using this
method medium-larger crystal growth was achieved (clear very light tan, almost colorless needles). The material was
vacuum filtered and placed in a vacuum desiccator to dry. Once dry the material was sized using brass sieves [12].

Melt-Cast Prills. To a 600 mL beaker was added 10 g of stock TNT, 300 mL of deionized water, and a 1” stir bar.
The contents of the beaker were heated with stirring to approximately 90 °C. Once all the TNT had melted, the heat
was turned off and the speed of stirring was increased to as fast as possible (without stir bar decoupling). The mixture
was allowed to cool to room temperature. This resulted in the formation roughly spherical shaped, orange-brown TNT
with a large particle size distribution. The material was vacuum filtered and placed in a vacuum desiccator to dry.
Once dry the material was sized using brass sieves.

SEM General Information. Scanning electron microscopy (SEM) was done by the M-7 High Explosive Science
and Technology group at Los Alamos National Laboratory (LANL). The SEM utilized was a JEOL 7900F field
emission microscope utilizing a Schottky Plus electron gun with an analytical resolution of sub 30 nm scale. SEM
images shown are those of lower electron detector (LED) emission capture.

SEM Procedures. The TNT samples received were in powder form, and 14 total samples were received and prepared
for imaging. Double sided carbon tape was utilized to adhere the sample to the stub. This tape also serves as an
electrical contact between the sample and the stub. Once enough material to take a representative data set is deposited
on the tape, the prepared specimen is placed into a gaseous glow discharge type sputter coater wherein an 8 nm sputter
coating of Au:Pd (80:20) was deposited on the sample. The metal coating helps to reduce charging and to improve
image quality relative to an uncoated sample.

050015-4
Drop Weight Impact Testing. Impact testing was performed using ERL Type 12 drop hammer equipment using a
2.5 kilogram weight, a 0.8 kilogram striker, an anvil and sound detection equipment. For each run 40 mg of material
was used with grit paper and was impacted using the drop weight from various heights. A go was defined as a 120
average decibel level from two sound detectors. The parameter reported is the DH50, which is defined as the height
from which 50% of the drops are a go. The DH50 values were calculated using the Neyer D-Optimal method for which
the higher the DH50 value corresponds to lower sensitivity to initiation.

Laser Diffraction Analysis. TNT powders were suspended in a solution of 1% Triton X-102 in water. The
suspension was inserted into an LS13 320 multi-wave length laser diffraction particle size analyzer (Beckman Coulter
Inc.), to a measured obscuration of 8-12%. All particles were sized using water as the fill medium, a 40% pump speed,
and a 60 second run duration. Each lot was sampled and sized in triplicate and reported as a standard mean derived
through the Fraunhofer theory optical model. The measurement of the light scattering pattern (collected from angles
of ~0° to 144°) is then translated into Heywood’s diameter or the equivalent circular area diameter (ECAD) which is
calculated according to ECAD = (4A/π)½, where A is the observed image area [13]. It is necessary to state that the
reported size distributions and distribution means are therefore spherical approximations regardless of their aspect
ratio [14]. The instrument was checked for accuracy prior to use with Coulter size standards LS Control G15 (nominal
15 µm garnet) and LS Control GB500 (nominal 500 µm glass beads).

CONCLUSIONS
In summary, 13 different samples of TNT were prepared from one lot historical sample of TNT, varying sizes of
crystals and melt-cast prills through preparation conditions and sieving. Impact testing was performed, and the DH50
values for most samples fall within error of each other, especially given the large standard deviations typically
associated with impact testing. The notable exception is the Lot TNT sample, which had significantly higher
sensitivity than the other samples. It is important to note that drop-weight impact tests are only a screening tool for
handling sensitivity, and interpretation of results can be difficult. Standard testing conditions vary from one laboratory
to another, so consistency among sample testing is critical. Additionally, the mechanisms of initiation for explosives
using impact testing depends on multiple phenomena that combine to provide enough heat for ignition somewhere in
the sample, through shear forces, bubble collapse/adiabatic compression of bubbles, and grit on grit interactions [7,15].
However, these preliminary studies indicate that particle size may be less important than expected in influencing
handling sensitivity of TNT, and further work is needed to understand the role that melting point has in these
experiments.

ACKNOWLEDGMENTS
The authors acknowledge Campaign 2 at Los Alamos National Laboratory for funding of this work. This work
was supported by the US Department of Energy through the Los Alamos National Laboratory. Los Alamos National
Laboratory is operated by Triad National Security, LLC, for the National Nuclear Security Administration of U.S.
Department of Energy (Contract No. 89233218CNA000001).

REFERENCES

1. V.W. Manner, T.W. Myers, M. J. Cawkwell, E. M. Kober, G.W. Brown, H. Tian, C. J. Snyder, and D. N. Preston,
“Understanding and manipulating the sensitivity of nitrate esters,” in Shock Compression of Condensed Matter-
2017, AIP Conference Proceedings 1979, Eds. R. Chau, T. C. Germann, and J. M. D. Lane (American Institute
of Physics, Melville, NY), 2018, p. 150027.
2. V. W. Manner, M. J. Cawkwell, E. M. Kober, T. W. Myers, G. W. Brown, H. Tian, C. J. Snyder, R. Perriot, D.
N. Preston, Chemical Science 9, 3649 – 3663 (2018).
3. V. W. Manner, M. J. Cawkwell, G. W. Brown, J. Yeager, L. Kay, M. Shorty, P. Leonard, B. C. Tappan, J. M.
Veauthier, “Examining the Effects of Crystal Structure and Bonding on Explosive Impact Sensitivity,” In
Proceedings of the 16th International Detonation Symposium; Cambridge, MD, 2018; pp. 1384–1393.
4. J. Liu, W. Jiang, Q. Yang, J. Song, G. Hao, F.-S. Li, Defense Technology 10, 184 – 189 (2014).
5. X. Song, F. Li, Defense Science Journal, 59, 37 – 42 (2009).
6. R. Matyas, J. Pachman, Primary Explosives, Springer, New York, NY (2013) p. 27.

050015-5
7. J. E. Balzer, J. E. Field, M. J. Gifford, W. G. Proud, S. M. Walley, Combustion and Flame, 130, 298–306 (2002).
8. V. W. Manner, B. C. Tappan, B. L. Scott, D. N. Preston, G. W. Brown, Crystal Growth & Design, 14, 6154–
6160 (2014).
9. N. Lease, L. M. Kay, G. W. Brown, D. E. Chavez, P. W. Leonard, D. Robbins, V. W. Manner, Crystal Growth
& Design, 19, 6708–6714 (2019).
10. J. E. Field, G. M. Swallowe, S.N. Heavens, Proc. R. Soc. Lond. A, 382, 231–244 (1982).
11. Urbanski, T.; Chemistry and Technology of Explosives Vol. 1; pg. 290.
12. Acken, M. F., and Woodbury, N. J.; Process for Crystallizing Organic Explosive Compounds; US Patent
2204059, 11 JUN 1940.
13. ISO Standard 13320-1 Particle Size Analysis–Laser Diffraction Methods: Part1. General Principle, International
Organization of Standardization, Geneva, Switzerland 1999.
14. R. Xu, O.A. Di Guida / Powder Technology 132 (2003) 145–153.
15. F. P. Bowden and A. D. Yoffe, Initiation and Growth of Explosion in Liquids and Solids, 1952, Cambridge at
the University Press, ch. 3.

050015-6