Skip to main content

Full text of "DTIC ADA601807: Template-Directed Crystallization of High Energy Materials"

See other formats


TECHNICAL REPORT 



Defense Threat Reduction Agency 
8725 John J. Kingman Road, MS 6201 
Fort Belvoir, VA 22060-6201 



DTRA-TR-14-30 


Template-Directed Crystallization 
of High Energy Materials 


Approved for public release; distribution is unlimited 


April 2014 


HDTRA1-08-1-0007 

Jennifer Swift 

Prepared by: 

Georgetown University 
Department of Chemistry 
37th and O Streets NW 
Washington, DC 20057 









DESTRUCTION NOTICE 


Destroy this report when it is no longer needed. 
Do not return to sender. 


PLEASE NOTIFY THE DEFENSE THREAT REDUCTION 
AGENCY. ATTN: DTRIAC/ J9STT, 8725 JOHN J. KINGMAN ROAD, 
MS-6201, FT BELVOIR, VA 22060-6201, IF YOUR ADDRESS 
IS INCORRECT, IF YOU WISH THAT IT BE DELETED FROM THE 
DISTRIBUTION LIST, OR IF THE ADDRESSEE IS NO 
LONGER EMPLOYED BY YOUR ORGANIZATION. 



REPORT DOCUMENTATION PAGE 


Form Approved 
OMB No. 0704-0188 


Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the 
data needed, and completing and reviewing this collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing 
this burden to Department of Defense, Washington Headquarters Services, Directorate for Information Operations and Reports (0704-0188), 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202- 
4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display a currently 
valid OMB control number. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS. 


1. REPORT DATE (DD-MM-YYYY) 2. REPORT TYPE 3. DATES COVERED (From - To) 

00-04-2014 Technical April 2014 


4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER 


Template-Directed Crystallization of High Energy Materials 


HDTRA1-08-1-0007 


5b. GRANT NUMBER 


6. AUTHOR(S) 

Jennifer Swift 


5c. PROGRAM ELEMENT NUMBER 


5d. PROJECT NUMBER 


5e. TASK NUMBER 


5f. WORK UNIT NUMBER 


7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 

Georgetown University 
Department of Chemistry 
37th and O Streeets NW 
Washington, DC 20057 


8. PERFORMING ORGANIZATION REPORT 
NUMBER 


9. SPONSORING / MONITORING AGENCY NAME(S) AND ADDRESS(ES) 

Defense Threat Reduction Agency 
8725 John J. Kingman Road STOP 6201 
Fort Belvoir, VA 22060 
PM/Suhithi Peiris 


12. DISTRIBUTION / AVAILABILITY STATEMENT 

Approved for public release; distribution is unlimited. 


10. SPONSOR/MONITOR’S ACRONYM(S) 

DTRA 


11. SPONSOR/MONITOR’S REPORT 
NUMBER(S) 

DTRA-TR-14-30 



14. ABSTRACT 


The objectives of this grant were to (a) examine the solution crystallization of RDX, HMX and CL-20 from a variety of solvents, 
withdetailed analysis of their phase, size, and morphological properties; (b) to prepare and fully characterize a library of 
gold-thiol and siloxane monolayers that can serve as crystal nucleation templates and (c) to assess the growth of RDX, HMX 
and CL-20 on these templates. 


15. SUBJECT TERMS 


High explosives, crystallization, RDX, CL-20, HMX 


16. SECURITY CLASSIFICATION OF: 


a. REPORT 

Unclassified 


b. ABSTRACT 

Unclassified 


c. THIS PAGE 

Unclassified 


17. LIMITATION 

18. NUMBER 

OF ABSTRACT 

OF PAGES 

SAR 

16 


19a. NAME OF RESPONSIBLE PERSON 

Suhithi Peiris 


19b. TELEPHONE NUMBER (include area 
code) 

703-767-4732 


Standard Form 298 (Rev. 8-98) 

Prescribed by ANSI Std. Z39.18 



































CONVERSION TABLE 

Conversion Factors for U.S. Customary to metric (SI) units of measurement. 


MULTIPLY- 
TO GET ◄- 


-> BY 


-> TO GET 


BY ◄- 


DIVIDE 


angstrom 

1.000 000 x E -10 

meters (m) 

atmosphere (normal) 

1.013 25 x E +2 

kilo pascal (kPa) 

bar 

1.000 000 x E +2 

kilo pascal (kPa) 

barn 

1.000 000 x E -28 

meter 2 (m 2 ) 

British thermal unit (thermochemical) 

1.054 350 x E +3 

joule (J) 

calorie (thermochemical) 

4.184 000 

joule (J) 

cal (thermochemical/cm 2 ) 

4.184 000 x E -2 

mega joule/m 2 (MJ/m 2 ) 

curie 

3.700 000 x E +1 

*giga bacquerel (GBq) 

degree (angle) 

1.745 329 x E -2 

radian (rad) 

degree Fahrenheit 

t k = (t°f + 459.67) /1.8 

degree kelvin (K) 

electron volt 

1.602 19 x E -19 

joule (J) 

erg 

1.000 000 x E -7 

joule (J) 

erg/second 

1.000 000 x E -7 

watt (W) 

foot 

3.048 000 x E -1 

meter (m) 

foot-pound-force 

1.355 818 

joule (J) 

gallon (U.S. liquid) 

3.785 412 x E -3 

meter 3 (m 3 ) 

inch 

2.540 000 x E -2 

meter (m) 

jerk 

1.000 000 x E +9 

joule (J) 

joule/kilogram (J/kg) radiation dose 



absorbed 

1.000 000 

Gray (Gy) 

kilotons 

4.183 

teraj oules 

kip (1000 lbf) 

4.448 222 x E +3 

newton (N) 

kip/inch 2 (ksi) 

6.894 757 x E +3 

kilo pascal (kPa) 

ktap 

1.000 000 x E +2 

newton-second/m 2 (N-s/m 2 ) 

micron 

1.000 000 x E -6 

meter (m) 

mil 

2.540 000 x E -5 

meter (m) 

mile (international) 

1.609 344 x E +3 

meter (m) 

ounce 

2.834 952 x E -2 

kilogram (kg) 

pound-force (lbs avoirdupois) 

4.448 222 

newton (N) 

pound-force inch 

1.129 848 x E -1 

newton-meter (N-m) 

pound-force/inch 

1.751 268 x E +2 

newton/meter (N/m) 

pound-force/foot 2 

4.788 026 x E -2 

kilo pascal (kPa) 

pound-force/inch 2 (psi) 

6.894 757 

kilo pascal (kPa) 

pound-mass (lbm avoirdupois) 

4.535 924 x E -1 

kilogram (kg) 

pound-mass-foot 2 (moment of inertia) 

4.214 011 x E -2 

kilogram-meter 2 (kg-m 2 ) 

pound-mass/foot 3 

1.601 846 x E +1 

kilogram-meter 2 (kg/m 3 ) 

rad (radiation dose absorbed) 

1.000 000 x E -2 

**Gray (Gy) 

roentgen 

2.579 760 x E -4 

coulomb/kilogram (C/kg) 

shake 

1.000 000 x E -8 

second (s) 

slug 

1.459 390 x E +1 

kilogram (kg) 

torr (mm Hg, 0° C) 

1.333 22 x E -1 

kilo pascal (kPa) 


*The bacquerel (Bq) is the SI unit of radioactivity; 1 Bq = 1 event/s. 
**The Gray (GY) is the SI unit of absorbed radiation. 








Template-Directed Crystallization of High Energy Materials 


Proposal: P2-06B-0022, Program: HDTRA1-08-1-0007 


Overview: 

The objectives of this grant were to (a) examine the solution crystallization of RDX, HMX and 
CL-20 from a variety of solvents, with detailed analysis of their phase, size, and morphological 
properties; (b) to prepare and fully characterize a library of gold-thiol and siloxane monolayers 
that can serve as crystal nucleation templates and (c) to assess the growth of RDX, HMX and CL- 
20 on these templates. 

To date, the work has resulted in 16 presentations and 2 published manuscripts. At least six 
additional manuscripts (full papers) are in various stages of completion. They should all be 
submitted within the next year and will acknowledge DTRA support. 

Accomplishments/New Findings: 

1. RDX 

(a) Slow evaporation growth of q-RDX from conventional solvents and co-solvents 

Crystal size and morphology are known to affect the sensitivity of explosive compounds. 
In formulating plastic bonded explosives (PBX) it is generally more desirable to use isometric or 
prismatic crystals rather than ones with highly anisotropic shapes (plates or needles) for packing 
efficiency reasons. However, PBX properties are also affected by the nature of the interactions at 
the crystal-binder interface. Thus, our first efforts focused on elucidatating how RDX crystal 
morphologies and sizes are related to the solvent conditions from which it is grown, and to put 
this in context with previous growth studies. 

Calculated morphologies (which do not account for solvent effects) predicted {111}, 
{200}, {020},{210} and {002} faces, consistent with previous reports. Experimentally, we 
found that RDX crystals exhibit highly variable morphologies when grown from the same 
solvent, and also when grown from different solvents. RDX crystals grown from acetone, DMF 
and cyclohexanone exhibited morphologies generally consistent with previous reports. Much of 
our efforts focused on examining growth from new solvents (THF, nitromethane, pyridine) as 
well as binary co-solvent mixtures. The latter were used in an effort to improve the solubility in 
less polar solvents. A summary of the morphologies, major faces and elongation axes observed 
appears in Table 1. 



Table 1. Summary of growth morphology, major faces and prism axis a-RDX grown from new solvent and 
co-solvent systems. 


Solvent 

Morphology 

Major Faces 
{hkl} 

Elongation 

Axis 

Reference 

Acetone 

Varies 

(001), (111) 

[010] 

This work 


Prism 

(120) 

[001] 

Elban et al. 1984 


Prism 

(102), (-210) 

[010] 

Elban et al. 1984 


Equant (Small) 

(001), (111) 

[010] 

Halfpenny et al. 1984 


Prismatic (large) 

(210) 

[001] 

Halfpenny et al. 1984 


Tabular (rare) 

(001), (210) 

[010] 

Halfpenny et al. 1984 

THF 

Needle 

Varies 

[010] 

This work 


Plate 

Varies 

[010] 

This work 

Nitromethane 

Prism 

(001), Varies 

Varies 

This work 

Pyridine 

Prism 

(001), (102), 
Varies 

[010] or [100] 

This work 

DMF 

Plate 

Varies 

Varies 

This work 


Prism 

(001) 

[010] or [100] 

Galdecki et al. 1984 


Prism 

(210), (111) 

[001] 

McDermott et al. 1971 

Cyclohexanone 

Large Plate 

(100) or (001) 

[010] 

This work 


Needle 

(100) 

[010] 

ter Horst et al. 1999, 
van der Heijden et al. 2004 


Plate 

(001) 

[010] 

Connick et al. 1969 

y-butyrolactone 

Prism 

(210), (111) 

[001] 

ter Horst et al. 1999, 2001 
van der Heijden et al. 2004 

DP 

Prism 

(100), (210), 

(111) 

[001] 

ter Horst et al. 1999, 
van der Heijden et al. 2004 

2:1 

Prism 

(001), (102), 

[100] 

This work 

Acetone:Nitrobenzene 

Plate 

(210), (111), 

(010) 

(001), (210) 

[010] 


2:1 Acetone: Benzene 

Varies 

Varies 

Varies 

This work 

2:1 Acetone:Pyridine 

Cubic Prism 
Triangular Prism 

(010), (111) 

Varies 

[100] 

[100] 

This work 

2:1 THF:Nitrobenzene 

Prism 

(102), (010), 

(210) 

[100] 

This work 

2:1 

THF: Cyclohexanone 

Prism 

Varies 

[010] or [001] 

This work 


To further analyze qualitative differences in RDX crystals grown from various solvents 
and co-solvents, a series of X-ray topography (XRT) experiments on whole RDX crystals was 
performed. Previous XRT work on RDX has employed large single crystals sliced along specific 
planes in order to quantify the magnitude, density and direction of dislocations. Our XRT work 
used whole single crystals in an effort to obtain a more wholistic view of the defect density as a 
function of growth solvent. A few early experiments were performed at Argonne National Lab in 
collaboration with David Black (Topometrix), using joint beam time awarded to Chad Stoltz 
(Indian Head), Kyle Ramos and Dan Hooks (LANL). This was a good learning opportunity, 
however, the decommisionning of the monochromatic beamline at Argonne in early 2009 
abruptly ended this line of inquiry, (note: this beamline is scheduled to come back online in late 





















2013) We redirected our efforts toward white beam topography experiments at the National 
Synchrotron Light Source (NSLS) at Brookhaven. It is worth noting that at the outset it was 
unclear whether white beam topography would even work on RDX samples, but we have 
demonstrated unambiguously that it does. There is still a large amount of analysis that needs to 
be done on the topography data that was obtained. This must be done in collaboration with 
topography experts. However, we can make some general observations which emphasize the 
importance of solvent effects. 

Comparison of acetone-grown plates and prisms (grown from the same batch) revealed 
that the former had more homogeneous diffraction. Darker diffraction bands indicative of grain 
boundaries were observed in prisms. Similarly, THF-grown plates and needles (again, grown 
from the same batch) showed differences in their topography, with needles generally showing 
fewer defects than the plates. Overall THF single crystals (both morphologies) had fewer defects 
than those grown from acetone. This suggests that while prisms are generally more desirable for 
PBX formulation, the defect densities of more anisotropic morphologies may offer advantages in 
terms of defects. RDX crystals grown from nitromethane, cyclohexanone and DMF showed that 
the defect structures varied greatly from sample to sample. DMF-grown crystals consistently 
showed lattice deformations (as opposed to inhomogeneities due to dislocations or lattice 
misalignments) which likely reflects solvent inclusion in the crystals. 

(b) Drop-cast evaporative growth on templates - characterization of [3-RDX 

RDX has five known polymorphic forms, though only the a and p phases are observable at 
room temperature and pressure. Many literature reports suggest that the metastable P-RDX 
polymorph is extremely rare due to the limited number of solvents from which it can be grown 
and its facile conversion to the more stable a-RDX form. Our efforts to control polymorphism by 
directed nucleation at designer surfaces revealed that P-RDX can be consistently obtained from a 
broad range of solvents (acetone, THF, nitromethane, DMSO) and that crystals grown this way 
remain stable over extended periods of time up to at least a year. This has enabled the most 
detailed analysis of their morphological and thermal properties to date. Profound differences in 
the behavior of a-RDX and P-RDX upon exposure to electron beam conditions were also 
observed. 

One pi drops of RDX in acetone, THF, nitromethane and DMSO drop cast onto plain glass 
and piranha cleaned glass slides resulted in fast evaporation and crystallization. Identification of 
a and P crystals by Raman spectroscopy revealed concomitant mixtures of a and P-RDX crystals 
were obtained from most of the drop cast experiments performed in acetone, THF and 
nitromethane, while crystallization from DMSO typically yielded exclusively one phase or the 
other in any given drop. Solvent choice also affected the morphology of the crystals obtained. 

In an effort to further explore concentration effects on the a:P distribution of RDX crystals 
formed, drop cast crystallization was performed from acetone at three different concentrations (5, 
10 and 30 mg/mL). Raman measurements of 50 separate crystals obtained under each set of 
conditions showed the overwhelming predominance of P deposits from low concentration drops 
(5-10 mg/mL), and more a crystals appearing from higher concentration drops (30 mg/mL). The 
switch to exclusively a-RDX at higher concentrations was ascribed to the larger number of 
crystals present coupled with convection currents in the evaporating solution which increase the 
probability of collisions and conversion to the more stable form. Other solvents showed similar 
trends, albeit with slightly different a:P ratios. 

Given the simplicity of the drop cast method, we also used it as a means to elucidate some of 
the thermal and material properties of P-RDX. RDX samples were prepared by drop cast 
crystallization from DMSO directly into aluminum DSC pans and subjected to phase analysis by 
Raman microscopy. Most samples initially identified as P-RDX exhibited an endothermic 



transition starting ~188°C and no subsequent transitions. Hot stage microscopy experiments 
confirmed this to be the melting transition. Samples of a- and P-RDX were also drop cast 
directly onto carbon grids, and SEM imaging revealed some interesting differences in the 
interactions of a and p single crystals with electromagnetic radiation. Focused beam irradiation at 
30 kV on the very tip of a p-RDX crystal grown from DMSO resulted in localized “bubbling” 
which we presume is decomposition. In contrast, spot irradiation on a-RDX crystals grown from 
nitromethane at 10 kV led to bubbling over the entire crystal. In a-RDX the transformation began 
in multiple spots and propagated throughout the crystal, whereas in the p form, the transformation 
began with cracking on the surface followed by localized bubbling in one spot on the crystal. 
From these observations it seems that the energy dissipation mechanisms in a and b are quite 
different and the localized effects and higher temperature required to effect decomposition in p, 
may have some advantages in select applications. 

(c) Oriented growth of q-RDX via slow evaporation on siloxane templates 

The template-directed nucleation of RDX was also examined on Au-S monolayers and 
siloxane monolayers under slow evaporation conditions. Our results show that the surface 
functionality and the growth solvent each play significant roles in determining the crystal 
morphology, nucleation density and surface orientation. In general, our efforts to control 
nucleation on gold-alkanethiol monolayers were not very successful, resulting in very high 
nucleation densities and large aggregates of crystals on the surface. Furture growth attempts on 
arylthiol monolayers which have 2D lattice spacings that are more compatible with RDX may be 
more successful. In one (as yet not reproduced) control experiment, we inadvertently grew single 
crystals of hydroxylammonium sulfate on a bare Au surface in our efforts to grow RDX from 
nitromethane. We still do not know if this compound was an impurity in the solvent or the RDX, 
but this unexpected result suggests a unique method to remove trace amounts of sulfurous 
impurities from a solution. Future efforts to follow up on this curious result may be pursued time 
permitting. 

Most of our efforts at template-directed nucleation focused on RDX growth on siloxane 
surfaces. In these slow evaporation studies, a greater emphasis is placed on understanding how 
the chemistry of the monolayer affects the nucleation density (and therefore crystal size), phase 
and morphology. RDX crystal growth on siloxane monolayers was attempted on alkyl and 
haloalky (fluoro, bromo, chloro and iodo), hydrogen bond donor and acceptor silanes (amino, 
cyano and isocyano) and phenyl derivative silanes (phenyl, pyridine and dinitrophenyl). Highly 
oriented crystal growth of a-RDX was generally seen only on siloxane surfaces bearing halogen 
groups and those capable of protonation. Detailed studies focused on 3-iodopropyl- (Si-3-I), 3- 
aminopropyl- (Si-3-NH 2 ) and 2,2-pyridylethyl- (Si-2-Pyr2) siloxanes and 3 different solvents 
(nitromethane, THF and 2:1 acetone:benzene). 

RDX was crystallized by slow evaporation from nitromethane (15 mg/ml), THF (10 
mg/ml) and 2:1 acetone/benzene (10 mg/ml) on Si-3-NH 2 , Si-2-Pyr2 and Si-3-I in fluorinated 
glass vials. In general, crystals grown on templates were smaller than comparable solution grown 
crystals. Prism-shaped crystals formed from nitromethane within 6-10 days and tended to form 
clusters. Needles crystallized from THF in 3-5 days on all SAMs and were of the same 
morphology as seen for solution grown crystals. RDX crystals grown from 2:1 acetone/benzene 
on Si-3-NH 2 , Si-2-Pyr2 and Si-3-1 surfaces within 1-2 days appeared to form needles. 
Morphologies from nitromethane (prisms) and THF (needles) are the same as in conventional 
solution growth. In contrast, the range of morphologies seen in 2:1 acetone:benzene was greatly 
reduced to just needles when surface-directed growth methods are used. 

Nucleation densities could not be obtained in nitromethane given the clustering observed, 
however, nucleation densities were systematically analyzed in THF and 2:1 acetone:benzene and 



are listed in Table 2. The narrow distribution of crystal sizes correlates with an obvious narrowing 
of the crystal size distribution, though we have not rigorously attempted to characterize particle 
sizes in this system. Although PXRD showed only diffraction lines corresponding to a-RDX, 
Raman microscopy of crystals grown from nitromethane on glass and Si-3-1 showed peaks in the 
ring-breathing region at 847 cm 1 (a-RDX) and 835 cm' 1 . The latter is consistent with P-RDX 
(but not HMX, which we considered as a potential impurity). The Raman data was collected on 
samples scraped from the siloxane. Given the mechanical transformation of P to a, we had 
assumed that scraping/grinding samples would have transformed any trace amounts of p, 
however, it appears that small amounts of P-phase impurities can survive this treatment. 

Table 2. Nucleation densities of a-RDX on differently functionalized siloxanes in nitromethane, THF and 
2:1 acetone:benzene. 


substrate 

nitromethane 

THF 

2:1 acetone/benzene 

piranha cleaned glass substrates 

clusters 

clusters 

5 ± 2 

Si-2-Pyr2 

clusters 

9 ± 1 

13 ± 6 

Si-3-NH 2 

clusters 

28 ±21 

3 ± 1 

Si-3-I 

clusters 

36 ± 21 

6 ± 1 


Crystals of RDX grown from nitromethane, THF and 2:1 acetone/benzene on different 
SAMs were analyzed by oriented PXRD while still attached to the substrates. Comparison with 
the calculated powder patterns for a-RDX (refcode: CTMTNA) and P-RDX (refcode: 
CTMTNA05) enabled identification of the Miller planes aligned at the siloxane/crystal interface. 
a-RDX has many systematically absent reflections in the 29 =10 - 20° range including: (110), 
(101), (Oil), (201), (120) and (012). None of these are typically dominant faces in a-RDX single 
crystals. Crystals grown from THF on all siloxanes showed a single intense peak at 20 = 17.92°- 
18.02° corresponding to the (102) plane. In contrast, crystals grown from nitromethane on all 
siloxanes were predominantly oriented on the (002) plane. 

In order to compare the degree of preferential orientation generated on the different 
surfaces, we attempted to quantify the ratios of the intensities for the four most intense peaks 
(111), (002), (021) and (102) relative to the simulated powder patterns. For crystals grown from 
THF, the degree of preferential orientation was 1.5 times on the Si-2-Pyr2 SAM and 1.8 times on 
the Si-3-1 SAM compared to crystal growth on glass substrates (which also show some 
orientation along (002)). No peaks corresponding to P-RDX appear in any template-directed THF 
growth experiments. For crystals grown from nitromethane, the (002) face was three times more 
intense on the Si-2-Pyr2 SAM, and two times more intense on the Si-3-I SAM compared to glass 
substrates. Glass substrates additionally showed a small amount of a-RDX oriented along (111) 
and (221) as well as P-RDX aligned along (111), (500) and (214). There was some evidence for 
small amounts of P-RDX growth on the Si-3-I and Si-2-Pyr2 surfaces, but only a-RDX was seen 
on Si-3-NH 2 . 

In an effort to establish a correlation between RDX sensitivity and the crystallization 
process, our studies suggest that defect densities as well as both chemical impurities and phase 
impurities deserve consideration. 




2. HMX 

(a) Solution growth from various solvents and co-solvents 


HMX has four known “polymorphic” forms: a-, [3- (the most dense and 
thermodynamically stable at RT), y-HMX (technically, a hemihydrate), and S-HMX (stable above 
160°C). Comparison of the intermolecular interactions between these conformational 
polymorphs using Hirshfeld surfaces provides some insight into their similarities and differences. 
For all polymorphs of HMX, the largest % of close contacts are seen between H*»*0 interaction, 
derived from C-H***0 interactions between neighboring HMX molecules in the lattices. The 
types of intermolecular interactions are most closely related for the p and y forms, while the a and 
8 forms are closely related. This is curious since the a, y, and 8 forms are all similar in 
conformation (chair-chair with the NCF groups pointing up) and the P form is different (chair 
with two NO 2 groups pointing up and two pointing down). However, the conformation of the p 
form means that the molecules can pack closely. 

In our experience, HMX is generally more difficult to crystallize than RDX. Slow 
evaporation growth of HMX at RT was examined from a variety of solvents and the resulting 
material examined by X-ray powder diffraction (PXRD). Results showed that growth from 
pyridine and cyclohexanone produced phase pure p-HMX. Growth from THF, acetone and 
nitromethane yielded concomitant mixtures of p-HMX and a-HMX. Growth from 2:1 
acetone:benzene yielded concomitant growth of the p-HMX and y-HMX forms. Crystals were 
often twinned. Some of these mixed phases may be due to phase transformations, since a y to P 
conversion upon grinding, and a solution-mediated a to P conversion have been previously 
reported. 

BFDH morphology calculations for the four HMX forms were done using Mercury CDS 
2.4. All three of the RT forms are predicted to have prismatic morphologies. Unlike RDX, a 
thorough examination of actual HMX growth morphologies has not been reported. We 
determined the growth morphologies of crystals grown from both single solvents (acetone, THF 
and nitromethane) and select co-solvents (2:1 acetone:benzene, 2:1 acetonemitrobenzene and 2:1 
acetone:DMSO). In general, HMX is less soluble than RDX, even in most polar solvents, but 
binary solvent mixtures help address this issue. A summary of growth morphologies of P-HMX 
appears in Table 3. 


Table 3. Summary of growth morphology, major faces, and prism axis for P-HMX grown from single and 
co-solvent systems. 


Solvent 

Morphology 

Major Faces {hkl} 

Elongation 

Axis 

Acetone 

Prism 

(0-11), (Oil), (110), (-101), (-110) 

[001] 

THF 

Needle 

(Oil), (0-21) 

[100] 

Nitromethane 

Prism 

(010), (110) 

[100] or [001] 


Plate 

(01-1) 

[100] 

2:1 acetone:benzene 

Elongated Prism 

(Oil), (110), (-110), (1-10) 

[100] 

2:1 acetonemitrobenzene 

Prism 

(011), (110), (-110), (1-10), 

(01-1), (0-11), (-101) 

[100] 

2:1 acetone:DMSO 

Prism 

Varies, (-1-10) 

[001] 


Monochromatic XRT of P-HMX crystals grown from nitromethane suggested that the 
majority of defects were due to grain boundaries. Comparison of the rocking curves one 2 
crystals - a prism and a plate, suggested that the former has a wider mosaic spread and more 
lattice misordering. This preliminary data indicates shape dependent defect densities, and is 
similar to our observations on RDX. White beam topography experiments were performed on 5 
P-HMX crystals grown from each of several solvents (THF and the three co-solvent mixtures 














listed in Table 3) in the hopes that differences in defect densities could be correlated with the 
growth solvents. Overall, the defect structure for P-HMX was dominated by lattice distortions 
most likely caused by solvent inclusions in the crystal lattice. This is not entirely surprising given 
that HMX is known to crystallize in dozens of solvate phases. There was some variation among 
crystals grown from the same solution as well as variation in crystals from different growth 
solvents, however, crystals grown from 2:1 acetone:benzene system generally exhibited the most 
uniform contrast. 

(b) Template directed growth of phase pure P-HMX on siloxane substrates 

Crystal growth of HMX on a variety of Au-S and siloxane monolayers using slow 
evaporation growth methods was performed. As in RDX studies, we generally had greater 
success with siloxanes. HMX crystal growth was attempted on alkyl and haloalky (fluoro, 
bromo, chloro and iodo), hydrogen bond donor and acceptor (amino, cyano and isocyano) and 
aromatic (phenyl, pyridine and dinitrophenyl) siloxane templates. Selective growth of P-HMX 
was observed on CN, NCO and NH 3 terminated surfaces and little growth was observed on 
surfaces terminated with F, Phenyl, CH 3 and others. One important lesson learned from these 
experiments was that experiments gave significantly different results in fluorinated glass vials, 
since fluorination suppresses nucleation on the walls of the growth container, resulting in much 
higher quality crystals on the siloxanes. 

Both solvent and template were found to affect the morphology of p-HMX. This was 
especially true in nitromethane and acetone, where crystals changed from prism to plates 
depending on the template. The P-HMX phase is usually observed exclusively but in a 
multiplicity of orientations. The exception to this observation is growth on the Si-3-CN surface, 
in which the elongated prisms grow nearly perpendicular from the surface with their a-axes 
emergent. Some of the needle-shaped crystals grew with the needle axis oriented perpendicular 
to the template (i.e. (1-10) plane parallel to the surface) while other needles appear to have fallen 
down and rest parallel to the template. Analysis of the (1-10) plane of P-HMX does not offer any 
clear molecular level reason why this orientation should be preferred. 

3. CL-20 

(a) Solution growth from various solvents and cosolvents - phase purity and morphology 
characterization 


There are five known polymorphs of CL-20: a (hydrate), p, y, s, and C , the first four of 
which can be obtained under ambient temperature and pressure conditions. The performance and 
stability of CL-20 can be affected by a number of its crystalline properties including the phase 
purity of the material, as well as the particle size, morphology and defect density of the individual 
crystallites. Slow evaporation crystallization in fluorinated glass vials was performed from 16 
different single solvent and co-solvent systems and the bulk crystalline material was assessed in 
terms of its phase purity by powder X-ray diffraction, hot stage microscopy and differential 
scanning calorimetry. These complementary methods confirmed that a concomitant mixture of 
polymorphs is typically obtained under most of the solution. Results are summarized in Table 4. 



Table 4. Summary of growth morphology, major faces, and prism axis for CL-20 grown from single and 
co-solvent systems. 


Solvent 

Morphology 

Phase 

Benzene 

prisms 

e,a,p (PXRD, DSC, XRD) 

Toluene 

prisms 

£, P (PXRD, XRD) 

1-Propanol 

prisms, 

plates 

£,P,a (PXRD, XRD) 

2-Propanol 


£ (PXRD) 

Nitromethane 

clumped prisms 

£, a, P, g (PXRD) 

Dichloromethane 

prisms 

£, P (PXRD) 

DMF 

red gel 

a,P (PXRD) 

9:1 2-Butanone:Benzene 

clumped prisms 

£,a (PXRD,XRD) 

3:1 2-Butanone:Benzene 

clumped prisms 

a, P (PXRD) 

1:1 2-Butanone:Benzene 

clumped prisms 

£, a (PXRD) 

3:1 2-Butanone:Toluene 

large prisms 

£,a (PXRD, XRD) 

1:1 2-Butanone: Toluene 

large prisms 

a,£ (XRD) 

3:1 2-Butanone:l-Propanol 

prisms, needles 

a,£,P (PXRD,XRD) 

3:1 Ethyl acetate:Benzene 

rounded prisms 

£,a (PXRD,XRD) 

5:4:1 Ethyl acetate: 1-Propanol: 
Benzene 

jagged prisms 

£,a,P (PXRD) 


Mixtures of a, |3 and s were observed from nitromethane, acetonitrile, acetone and 1 - 
propanol, while typically only s crystals were obtained from toluene and 2-propanol. Solvents 
with lower boiling points yielded smaller crystal sizes (presumably from less crystallization time) 
but had no observable effect on the phase. For binary solvents, one solvent with average CL-20 
solubility (benzene, toluene, 1 -propanol, 2-propanol) is paired with a solvent of high solubility 
(ethyl acetate, 2-butanone) in varying compositions. This allowed for more concentrated solutions 
without the addition of heat. A greater variety of forms (e.g. plates, prisms, needles) are observed 
in the co-solvent systems relative to the single solvent growth which typically yields prisms. 

Morphological characterization of numerous individual crystals was also performed using 
single crystal X-ray goniometry, and the Miller indices compared against calculated BFDH 
morphologies. All solutions yielded heterogeneous crystal sizes and 1-propanol exhibited the 
widest variety of morphologies. BFDH calculations for s-CL-20 suggested {002}, {Oil}, {110} 
and {101} to be important families of surfaces governing crystal morphology. Experimental 
findings show these families were typically the largest observed for the majority of crystals 
grown, with (10-1) and (110) often appearing as the large faces seen for e-plates. Growth in 3:1 
l-Propanol:2-Butanone also yielded well-formed P prisms and needles large enough to be 
isolated for indexing. Examination of the packing interactions via Hirshfeld surface analysis 
showed great similarities among the different polymorphs, which is presumably a contributing 
factor in their concomitant crystallization. 

(b) CL-20 growth on siloxane monolayers - phase and particle size control 

CL-20 crystallization on siloxane monolayers was performed resulting in two very 
interesting trends. Large differences in the nucleation densities were observed, though the trends 
were solvent dependent. For example, phenyl and Cl terminated surfaces nucleated 2-6 X fewer 
crystals in 1- propanol compared benzene, but CH 3 and NH 2 terminated surfaces nucleated 2-3X 
more crystals in 1-propanol than benzene. A nucleation density study performed in benzene 
revealed only s growth on multiple siloxanes surfaces. For any given siloxane, increasing the 
concentration resulted in increased nucleation density rather than increased crystal size, 





















suggesting that siloxane-directed nucleation can be used to generate crystalline materials with 
narrow particle size distributions. Furthermore, the crystals nucleated on these surfaces show 
strong preferred orientations (much stronger than either RDX or HMX). 

Many template/solvent combinations yield exclusively s in multiple orientations. We 
have also identified conditions which give phase pure a or p. Analysis of the siloxane/crystal 
interfaces is ongoing, though the preference for a given set of conditions to yield a metastable 
phase and/or highly oriented e may or may not provide clues for the molecular-level origins of the 
surface-directing effects. Computational modeling efforts would we welcome on this front, and 
toward that end we are hoping to start a funded collaboration with computational chemists at the 
University of Missouri who can help on this effort. 


Personnel Supported: 

Jennifer A. Swift, faculty member responsible for overseeing project as a whole 
Pranoti Navare PhD (postdoctoral associate, 6/12 - 5/13) 

Jessica Urbelis (graduate student, 1/08 - 5/13) 
liana Goldberg PhD (graduate student, 12/07 - 10/11) 

Christina Capacci-Daniel PhD (graduate student, 12/07 - 9/09) 

Cameron Mohammadi (undergraduate, 1/12 - 5/13) 

Adam Hoy (undergraduate, 10/09 - 12/10) 

Brian Fochtman (undergraduate, 9/08 - 1/10) 

Lindsey Roeker (undergraduate, 3/08 - 5/10) 

Aliza Cruz (undergraduate, 9/08 - 1/09) 

Publications*: 

Jessica H. Urbelis and Jennifer A. Swift, “Solvent Effects on the Growth Morphology and Phase 
Purity of CL-20,” in preparation 

Pranoti S. Navare, liana G. Goldberg and Jennifer A. Swift, “Oriented Growth of RDX on 2-D 
Templates,” in preparation 

liana G. Goldberg and Jennifer A. Swift, “New Insights into the Metastable P Form of RDX,” 
Crystal Growth & Design , 12 (2), 1040-1045 (2012) 

Christina Capacci-Daniel, Karen J. Gaskell and Jennifer A. Swift, "Nucleation and Growth of 
Metastable Polymorphs on Siloxane Monolayer Templates," Crystal Growth & Design , 10, 2, 
952-962 (2010). 

*A minimum of 5 additional manuscripts acknowledging DTRA support are forthcoming 
including analysis of our studies on: (1) Solvent effects on the growth morphology of RDX; (2) 
Solvent effects on phase purity and morphology of HMX; (3) Template-directed growth of HMX 
on siloxane monolayers, (4) Controlling CL-20 polymorphism and particle size by growth on 
siloxane monolayer templates; (5) Changes in the surface chemistry of silicate glass upon 
exposure to different solvents and temperatures. 



Interactions / Transitions: 

(a) Conference participation and invited talks 

“Crystallization of Energetic Materials on Monolayer Templates,” University of Missouri - 
Columbia (Columbia, Missouri) to be given Sep 2013 (rescheduledfrom Feb 2012 due to 
weather-related airline cancellation), (invited lecture) 

“Crystallization of Energetic Materials on Monolayer Templates,” Los Alamos National Lab 
(Columbia, Missouri) to be given Jul 2013 (rescheduled from Feb 2012 due to conflict with 
awarding of beam time), (invited lecture) 

“Crystallization of Energetic Materials on Monolayer Templates” Naval Air Weapons Station 
China Lake (China Lake, CA) Dec 2012. (invited lecture) 

“Crystallization of RDX on 2D Templates: Surface and Solvent Effects,” Pranoti Navare , liana G. 
Goldberg and Jennifer A. Swift, 245 th ACS National Meeting (New Orleans, LA) Apr 2013 
(poster). 

“Crystallization of CL-20 on Monolayer Surfaces,” Jessica Urbelis and Jennifer A. Swift, ACS 
Midwest Area Regional Meeting (UMBC, MD) May 2012. (poster) 

“RDX Polymorphism Revisited,” liana G. Goldberg and Jennifer A. Swift , IS' 1 ' International 
Symposium on Industrial Crystallization (ISIC18), (Zurich, Switzerland) Sept 2011 (poster). 

“Crystallization of CL-20 on Monolayer Surfaces,” Jessica Urbelis and Jennifer A. Swift, 
Defense Threat Reduction Agency’s Basic Research Program for Countering Weapons of Mass 
Destruction Technical Review, (Springfield, VA) Jul 2011. (poster) 

“Crystal Growth of HMX on Self-Assembled Monolayer Templates,” liana Goldberg and 
Jennifer A. Swift, Defense Threat Reduction Agency’s Basic Research Program for Countering 
Weapons of Mass Destruction Technical Review, (Springfield, VA) Jul 2011. (poster) 

“Polymorphism of the Secondary Explosive RDX Revisited,” liana G. Goldberg and Jennifer A. 
Swift, AC A Annual Meeting (New Orleans, LA) May 2011. (seminar) 

“Polymorphism of the Secondary Explosive RDX Revisited,” liana G. Goldberg and Jennifer A. 
Swift, ACS Midwest Area Regional Meeting (College Park, MD) May 2011. (poster) 

“Crystallization of the Energetic Materials RDX and HMX: Morphology and Structure 
Properties,” liana G. Goldberg and Jennifer A. Swift, Materials Research Society Annual Fall 
Meeting (Boston, MA) Nov 2010. (poster) 

“RDX Crystallization in Different Environments: Solvent and Template Effects” liana G. 
Goldberg and Jennifer A. Swift, Defense Threat Reduction Agency’s Basic Research Program 
for Countering Weapons of Mass Destruction Technical Review, (Springfield, VA) Aug 2010. 
(poster) (*Best Poster Award*) 

“RDX Crystallization in Different Environments: Solvent and Template Effects” liana G. 
Goldberg and Jennifer A. Swift, Gordon Research Conference Energetic Materials, (Tilton, NH) 
Jun 2010. (poster) 



“Crystal Growth of Polymorphic Energetic Materials Using 2D Self-Assembled Monolayers” 
liana G. Goldberg, David Black and Jennifer A. Swift, 238 th ACS National Meeting, 
(Washington, DC) Aug 2009. (poster) 

“Directed Crystallization on 2D Monolayer Templates,” liana G. Goldberg , Lindsey E. Roeker, 
David Black, and Jennifer A. Swift, ACA Annual Meeting, (Toronto, Ontario) Jul 2009 (seminar) 

“Crystal Growth of Polymorphic Energetic Materials,” liana G. Goldberg and Jennifer A. Swift, 
Defense Threat Reduction Agency’s Basic Research Program for Countering Weapons of Mass 
Destruction Technical Review, (Springfield, VA) Oct 2009. (poster) 

“Characterization of Polymorphic Compounds,” liana G. Goldberg and Jennifer A. Swift, ACA 
Annual Meeting, (Knoxville, TN) May 2008. (poster) (*Pauling Poster Award*) 

“Characterization of Polymorphic Compounds,” liana G. Goldberg and Jennifer A. Swift, GRC 
Energetic Materials, (Tilton, NH) Jun 2008. (poster) 


(b) Consultative and advisory functions to other labs & agencies & other DoD laboratories 
NSF - Proposal reviewer, CHE and DMR divisions 

(c) Transitions 
Nothing yet to report 

New discoveries, inventions and patent disclosures 

In discussions with university lawyers about patenting the phase selective growth of CL20 
using our template methods. 



DISTRIBUTION LIST 
DTRA-TR-14-30 


DEPARTMENT OF DEFENSE 


DEFENSE THREAT REDUCTION 
AGENCY 

8725 JOHN J. KINGMAN ROAD 
STOP 6201 

FORT BELVOIR ,VA 22060 
ATTN: S. PEIRIS 

DEFENSE TECHNICAL 
INFORMATION CENTER 
8725 JOHN J. KINGMAN ROAD, 
SUITE 0944 

FT. BELVOIR, VA 22060-6201 
ATTN: DTIC/OCA 

DEPARTMENT OF DEFENSE 
CONTRACTORS 

EXELIS, INC. 

1680 TEXAS STREET, SE 
KIRTLAND AFB, NM 87117-5669 
ATTN: DTRIAC 

GEORGETOWN UNIVERSITY 
DEPARTMENT OF CHEMISTRY 
37 th AND O STREETS NW 
WASHINGTON, DC 20057 
ATTN: J. SWIFT 


DL-1