NOLTR 69-67

VAPOR PRESSURES AND HEATS OFSUBLIMATION OF SOME HIGH MELTINGORGANIC EXPLOSIVES

00

ByJerome M. RosenCharles Dic',ins(,n

N ot 16 APRIL 1969

UNITED STATES NAVAL OiRDNANCE LABORATORY, WHITE OAK, MARYLAND

ATTENTION

0 This documenm has been Approved for

Z public release and sale, its distributionis unlimited.

NOLTR 69-67

VAPOR PRESSURES AND HEATS OF SUBLIMATION OF SOME

HIGH MELTING ORGANIC EXPLOSIVES

By

Jerome M. Rosen and Charles Dickinson

ABSTRACT: The Langmuir method was used to determine the vapor pressuresand heats of sublimation of several high melting organic explosives.The teMperatures at which each compound would have a vapor pressureof lO- torr are: (1) cyclotrimethylenetrinitramine (RDX), 48.0°C;(2) cyclotetramethylenetetranitramine, beta polymorph, (8-HMX),121.90C; (3) 2,4,6-trinitroaniline (TNA), 44.0°C (4) 1 3-diami.no-2,4,6-.trinitrobenzene, form I polymorph, (DATB-I5, 79.86 C4 (5) 1,3,5-triamino-2,4,6-trinitrobenzene (TATB), 131.4 0 C; (6) 2,2' ,4,4',6,6'-hexanitrostilbene (HNS), 170.6 0 C. Within a series those compoundswhich form the strongest intermolecularhydrogen-bonds have the highestheats of sublimation.

Carl Boyars, ChiefAdvanced Chemistry Division

CHEMISTRY RESEARCH DEPARTMENTU. S. NAVAL ORDNANCE LABORATORY

White Oak, Silver Spring, Maryland

ii

NOLTH 69-67 16 April 1969

VAPOR PRESSURES AND HEATS OF SUBLIMATION OF SOME HIGH MELTINGORGANIC EXPLOSIVES

Explosives have been used to perform a variety of functions in space

and military systems. Information contained in this report makes itpossible for design engineers to select the appropriate explosivefor each device and mission. The very low vapor pressure of HNS hascontributed to its selection for many applications in the Apolloprogram. This work was done under Task Numbers ORD-332-001/092-1/UP17-354-301, Explosives Characterization and MAT 03L O00/ZR 011 01 01,Explosives Chemistry.

E. F. SCHREITERCaptain, USNCommander

ALBERT LIOTBODY-By direction

U

S]

I

IIa

NOLTR 69-67

TABLE OF CONTENTS

Page

INTRODUCTION . . . . . . . . . . . . . . . 1

EXPERIMENTAL . . . . . . . . . . . . . . . . . . . . . . . . 1

RESULTS AND DISCUSSION ...... ......... .... 8

ACKNOWLEDGMENT . . . . . . . . . . . . . . . . . . . . . 20

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . 21

ILLUSTRATIO,'IS

Figure Title Page

1 Vapor Pressure Apparatus . . . . . . . . . . . 32 Sample Holders . . .. .. ..... . . . .. . 43 Temperature Correction Curve . . . . . . . . . . . 6

4 Vapor Pressure of RDX . . . . . . . . . . . . . . . 12

5 Vapor Pressure of TNA . . . . . . . . . . . . .*. . 13

6 Vapor Pressure of DATB-I . . . . . . . . . . . .a. 14

7 Vapor Pressure of Beta HMX . . . . . . . . . . . . 15

8 Vapor Pressure of TATB . . . . ....... .. 16

9 Vapor Pressure of HNS . . . . . * . * . . * * . . 17

TABLES

Table Title ?age

1 Temperature of Reflux Liquids . . . . . . . . . . . 5

2 Results and Data . . . . . . . . . . . . . . . . . 9

3 Heats of Sublimation . . . ..*. . . . . . . . ... 19

ii'

NOLTR 69-67

VAPOR PRESSUIES AND HEATS OF SUBLIMATION OF SOMEHIGH MELTING ORGANIC EXPLOSIVES

INTRODUCTION

Measurements of the vapor pressure of high energy compounds areof both theoretical and practical importance. From the practical

viewpoint, one is able to compute a rate of recession of a freesurface of an explosive heated in a hard vacuum (equivalent to outer

space), as a function of temperature. This type of information is of

value to the design engineer concerned with space applications ofexplosive devices, particularly those involving long exposure timesat elevated temperatures. If theee is a potential problem arisingfrom losses of material in an explosive device by sublimation, the

engineer is able to estimate the magnitude of the problem by cal-culating recession rates.

The vapor pressure and heat of sublimation data are of theoreti-cal importance in computing bond energies. For bond energy estimates,the heat of formation of the vapor is required. This is the sum ofthe heat of formation of the solid and the heat of sublimation. Be-cause the heat of sublimation has been difficult to determine and

thus generally not available, it has been estimated by many in-P vestigators. The direct determination of the heat of sublimation

removes a source of uncertainty in calculating bond energies.

EXPERIMENTAL

The Langmuir method was used for the determination of vaporpressures from measurements of the rates of sublimation at constant

9,4temperature9. Vapor pressures were computed from the observedweight logs of the sample heated in a vacuum. The expression used

was:F(torr) = 17.14 x G x (T/M)i, where

G is the weight loss in gm/cm2 /sec,T is the absolute temperatureM is the molecular weight.

NOLTR 69-67

In the calculations a condensation coefficient of 1 is assumed. This,

we feel, is ju tified by the good agreement between the results ob-

tained from the Langmuir and Knudsen method results for RDX.

Samples were heated at a constant temperature in a glass vacuum

apparatus, Figure 1. After centering the sample holder in the

jacketed section, the section was sealed and the system evacuated to

at least 5x10- 7 torr. Two types of sample holders were used,

Figure 2. One consisted of a metal plate with machined, circular

depressions. The other type was constructed by cutting a glass

sample vial about 8 mm from the closed end. The .ample was heated

by passing the vapors of a refluxing liquid through the Jacket of

the sample section. Two 500 ml round-bottomed flasks heated by

electric mantles and located near the ends of the jacketed regicn

were used for the reflux liquid, and the vapors were condensed by a

single water-cooled condenser located as indicated. The Jacketed

section was insulated with eight layers of aluminum foil.

Constant temperatures were provided by a series of liquics,

Table 1, maintained at the reflux point. The boiling points of the

liquids were determi.ned in the usual manner. However, the tempera-

ture at the sample is lower than the boiling point of the refluxirg

liquid. For example, in the case of water this diff-rence was 2.1 0 C.

A temperature correction then, was applied for all the liquids used.

Temrperatures irride the vacuum apparatus at the sample position

were dee7%•iAned for five of the liquids at reflux. This was done by

Qealing a ohrcmel-alumel thermocouple inside the apparatus. From

the measurements of these five liquids a temperature correction

curve, Figure 3, was constructed from the differences between the

temperature inside the vacuum apparatus and the temperature of the

refluxing liquid. Temperature corrections for all liquids were

taken from this curve.

We feel that the uncertainty in the timperature represents the

largest source of experiment'O error. This uncertainty is estimated

to be about 0.2 0 C to 0.30C' _ ,. e temperatu es below 1500 C for

heating periods less than 1; hours. At higher temperatures and

2

NOLTR 69-67

pd

U5-

CLJ

U-S

OC-

LA4

NOLTR 69-67

AC -F!:G. 2 -4AMPI.LE HOLDERS

NOLTR 69-67

TABLE 1. Temperature of Reflux Liquids.

Boiling Temperature InLiquid Temperature* Vacuum Apparatus

Methyl acetate 57.40C 55.7 0 CMethanol 64.4 62.6**Benzene 80.1 78.2n-Propanol-water azeotrope 87.3 85.3**

Water 99.8 97.7*Toluene 110.3 108.1n-Butanol 117.8 115.6Cblorobenzene 131.8 129.3m-Xylene 138.8 136.2Anisole 152.7 150.0Mesitylene 164.4 161.4"*Phenetole 169.4 166.3o-Dichlorobenzene 180.5 177.3Benzonitrile 189.3 185.8Nitrobenzene 210.3 206.3*

• At 755 torr atmospheric pressure

•* Temperatures deterrlned by thernmocouple measurementswith external p-ressure of 755 torr

h.5

NO LTR 69-67

Lu

UL Z0 >-

LL<L Z .LZLCLU

'u LU 0II 1- U

PL)

LUj

0a-

U-

0

0 0C0 0 C0

6

NOLTR 69-67

longer times the estimated error is slightly greater, perhaps O.50 C.Barometric pressure fluctuations affect the boiling points of theheating liquids and this is likely to be a source of some errorduring the longer runs.

About 40 to 50 minutes are required for the sample to reach itsmaximum temperature after beginning to heat the reflux liquid. Thistime interval was determined by following the temperature increase

at the sample position as a function of time. To determine time

intervals for the experimental r'uns a time of 40 minutes after the

initiation of heating was used as time zero in every case. At the

completion of the heating period the jacketed section was cooledwith an air gun for about 15 minutes, after first removing a part

of the aluminum fuil insulation. The error in the measurement of

the heating interval is probably about 0.2 hour.

The vacuum system consists of an eight liter per second Varian

Vac Ion pump and a Varian titanium sublimation pump connected to a

stainless steel manifold. The manifold is a cross fitting with four

stainless steel, high vacuum valves in the arms of the cross. One

valve isolates the pumps from the system, a second valve isolates a

fore pump. The remaining two valves connect two glass, jacketed

units to the vacuum system through glass to metal seals. All glass

parts are made of 1.5 inch I.D. borosilicate glass. Each unit maybe operated independently.

All of the samples were carefully purified by repeated re-crystallization and thoroughly dried to remove residual solvent. In

general, these materials represent the highest quality available for

study. The purity of the samples was tested by at least one of the

following techniques in each case: thin layer chromatography,

melting poi.t, or gas chromatography. The compounds in this work

have all been studied at temperatures far higher than any used inthe determination of vapor pressures. At these elevated temperatures

no decomposition was detected as determined by gas evolution

techniques 1 1 and thin layer chromatography7 .

7

NOLTR 69-67

HMX has four well characterized polymorphs. P-HMX is stable up

to 10200, although it does not undergo a solid-solid transition until

well above 1500C. Thus, all of the measurements on P-HMX except the

one at 97.6 0 C are in the temperature range where this crystal modi-

fication of HMX is unstable. It was necessary to heat P-HMX for

forty days at 97.6 0 C to obtain a measurable loss in weight. At the

completion of this run, the apparatus was rinsed with acetone. All

of the sublimed material was collected and identified as P-HMX by

refractive index measurements, indicating no decomposition had oc--

curred. Upon completion of the run at the highest temperature,

129.30C, the sample wad examined by X-Ray diffraction and found tostill be the p polymorph.

DATB has two well characterized polymorphs, the I-II transition

occurring at 2170C. In contrast to HMX, DATB easily undergoes this

transition even in the solid phase, which may be explained by the

rather higher vapor pressure of DATB. However, the highest tempera-

ture used for DATB-I in these vapor pressure measurements (1080C) is

over 1000 lower than the transition point. At these relatively low

temperatures no detectable amount of I÷II conversion has ever been

found.

RESULTS AND DISCUSSION

The vapor pressures of the compounds at the temperatures of the

experiments are reported in Table 2, along with a summary of the

experimental data. The least squares fits of the vapor pressures to

the equation:

logloP = B/(4.576 x T°K) + A

are given in Figures 4 through 9. The line constants A and B, B

representing the heat of sublimation, are also given in Table 2.

The vapor pressure of RDX is in good agreement with the data

obtained by the National Bureau of Standards (NBS) using a Knudsen

cell 1 0, The NBS sample, supplied by this laboratory, was a part of

the same sample used in this work. Vapor pressure measurements on

RDX by Edwards using a Knudsen cell are somewhat higher, although

his data at 131 0 C and 1380 C would appear to be in reasonableagreemcnt with the other data5 .

8

NOLTR 69-67

C/3 0 0

KI rN cQ

0O0

H 4z Cu

Lm C H H - kCC K\ noNON

":NLnQ CIA nnnmmu OtD 0rli 0t0 n0M\ -o tA ntC

Cd a 000000000000 0000000000004-3) ' cz]o OC cr0 0 0 0 0 0 N CN Coo 0 0coao'o'..o1(d C 0 0 0 NGOW\tý0 ID H H 0 0 N N N NQQ'I H t-t--

0 . . . .S . .S % "S * ftO ' -1 05 -ý "S . - ft '

a)'UCjN_1 -T-- -tl nCjNO MI CUCU r' Cu-K'N \t- tt- Q N CujCfl 4rt4HH -i 1- Cuj j K K K\\

Cl)

02 0M

H, 0\LC\'O Ltý 0 0u t-O~COu t- CC) ý00k Lr\-CD H4 00 H nCjko

cm F-i H .- N n 0 *\ *n H H H *- H* L

w 0O 0 40-1V Nr co t--N0C00Pý4-3 Lf) C\J (\JH'Or N\ -tDM Or- -

t-- Ký' N \C A~C O W t f co (:(D -T Cuj LC\ Lrý 4 0 CO SýCO nc\o Ut-- L Kl\"() 0H I'D '.0 .D Cuj Cu

OH- t-ý'~Dt-- Q Q'~

0

C)ý

N. LrxC\ nýO1,DucD[I-t.\-r- C J --rŽ--Dk t-t- - 'DU

sx

0p

9

NoLTR 69-67

0 0 0ON r-HcoNH 0

H

H ~HH

r~~~\~ N ~ n\ Ml\ N \t\ (

0 %D C0 CM \J 0 0 9 0 0 00 S 0 8 S 5 0 004C~i -- c CMn l"O-z 0 0 04004o H~ H 040 Go HI H 0 4 t

0 rnk.a~o00O * Lr\ uk O ' : tri - -t-t-'coc

43

0

tc*: 01 M 0- K' 0 S-k ON a\S 0 5 5 5 0 H n n 0 0- kDO 0 0 0-

N 0 4Ir- r H HH' HHH04N0--r44

E44

C) K\ 0 0 0 P- 0 0 5- n nz 0 0 So t0 V5 0\ 5 -z t- N 0 SS 0

00004 OQ'C ; 3 ýCý'ýC C 4 00000 00 H ý H ý 0 ý O0C\J 04 i

ko 0~ n n 5k 5o 0- \0 0~ %5 51 \0 n n Sn N 0 0 0----

H0Ln n rt- t- 0 - ccOL L)Ma (NC~ ~00 0 HH\CQ -

H HHHH IH H r- r4r- rIH H rl rfH rIHH

10

NOLTR 69-67

0H0

(A

C~O~D coJC~

\ýo co 0 co

0%

U~ 0 Ln La

0~ 0 j4-0

-I--- Ka~c' iC\

kD \0 S 0 S

H\ Uq H- C~1-

NOLTR 69-67

0 AD

-I-

p < -0I Lu

ruj

0

04 U

I 0 0 0? 0 0

NOLTR 69-67

0'

a..

u.j -J z

UNU

0 u.j

uJ

UAJ 0

C4U,

C? W')0

(noi) ýnsS3 S~dV

-V ~i13

NO LTR 69-67

(1 • I e•I-

I IIe*.d

144

C4

I. ICD

IýOl WsLJJ~V

14 '

NOLTR 69-67

10-68

/

0 a -

2- /O Arts =41.9AK CAL/MOLE

10 - - _ _ _ _ _ _ _ _ _ _ __ _ _ _ _ _ _ _

-0-9

2.7 2.6 2.5 2.4TEMPERATURE 1000

T,*K

FIG. 7 VAPOR PRESSURE Ff.3--4MX

iI

NO LTR 69-67

-j

C-4

ae

C

~-ro

A 0

w0

0o

ýoo ) 3dlSS3d ýOdVA

16

NOLTR 69-67

10-5

6-

4 -

2-

10 -6 .

-6

0t.-

LnD

LUJ

O AHs 43.0KCAL/MOLE

10-7

6

4

10-82.2 1 2.0

TEMPERATURE :1000

riG. 9 v',APO, PRESSURE OF HNS

NOLTR 69-67

Ignoring any effects caused by not correcting to a standardtemperature, several interesting observations may be made on theheats of sublimation and the properties of the compounds. Thus thedifference between the molar heats of sublimation of RDX and P-HMX,

10.78 kcal/mole,, apparently is largely a function of molecularweight. This ia indicated by the specific heats of sublimation,140.0 and 141.4 cal/,.m, respectively, and agrezz with the known

structures of ý-H!C1 and RDX6 which indicate no marked difference inintermolecular forces.

On the other hand, the trend in the molar heats of sublimationin the eeries of trin' cro-aromatic amines persists even when thespecific heats are considered, Table 3. This tread agrees, at leastin a qualitative sense, with what we know about the structures ofthese solids based on published crystal structure determinations.The structures of DATB-I 8 and TATB2 are quite different. The formeris hydrcgen bonded into chains with two intermolecular hydrogenbonds per molecule within a chain and only weak interactions betweenthe chains. TATB, however, is hydrogen bonded into a layered struc-

tz - with six strong intermolecular hydrogen bonds per molecule.Although the structure of TNA has not been reported, the structuresof other nitro anilines such as p-nitroaniline1 2 and 2,3,4,6-tetra-

3-nitroaniline have been determined and these compounds exhibithydrogen bonding schemes similar to one another and weaker than thatfound either in DATB-I or TATB, ba3ed on the intermolecular separa-tion between amino and nitro functional groups. If the structure ofTNA is comparable to these compounds, at least with respect to inter-molecular hydrogen bonding, then the intermolecular forces due tothis cause in TNA should be smaller than in DATB-I which in turnshould have smaller forces than TATB. Thus ignoring other factorswhich affect the heats of sublimation (polarity, dispersion, tempera-ture, etc.), it can be seen that within this series of polynitro-aromatic amines the trends in intermolecular hydrogen bonding andheats of sublimation parallel one another in a qualitative sense.A more exact discussion would require inclusion of the factors ig-nored here as well as some method for esat*mating the strengths ofhydrogen bonding systems that is more meaningful than internuclearseparation. 18

NOLTR 69-67

04 M

~ý4z 9400(

lo t. - -- oLc

vi "u- *O4)0

4) ~ 0

4.) Cd V

00

H 4-41 (LCd 4-) 0 H

40 . . 0 000b \ Q 00r K. 0 m4) 0

-H(L C)j 004 -J U4-) H N N ON)7-4-4,0 .3 .9 . d*O

0.4-OXV4-)

04-) r-i04) H4-)

C.) -H) 0

C; V- H 4)r- ri d0

CJ43 Q

04) W

0 aH-ri 0344. H H * '-H 0

Cd4-) tdO

00

4)

H ~ ~ ~ ~ .- (- MHt-J43-I 0

0-4 :% ... Xr O00 V2 rr .I -n n

00 *

r-4-9

NOLTR 69-67

ACKNOWLEDGMENT

The authors are deeply indebted to Mr. Joshua D. Upton and

Miss Eleonore Kayser for their assistance in the performance of the

experiments and the preparation of this paper.

20

NOLTR 69-67

REFERENCES

1. Cady, H. H. Larson, A. C. and Cromer, D. T., Acta Cryst.,

16 617(1963).

2. Cady, H. H. and Larson, A. C,, Acta Cryst., 18, 485(1965).

3. Dickinson, C., Stewart, J. M. and Holden, J. R., Acta Cryst.,21, 663(1966).

4. Dushman, S. and Lafferty, J. M., "Scientific Foundations ofVacuum Technique", 2nd Efition, 1962, John Wiley & Sons.

5. Edwards, G., Trans. Faraday Soc., ý_, 152(1953).

6. Harris, P. M., AFOSR TR-59-165, available to qualified re-questers from DDC, Cameron Station, Alexandria, Va.

7. Hoffsommer, J. C., Private Communication.

8. Holden, J. R., Acta Cryst., 22, 545(1967).

9. Langmuir, I., Phys. Rev., 2, 329(1913).

10. Miller, R. G. and Prosen, E. J., Private Communication, 1953.

11. Rosen, A. H. and Simmons, H. T., NAVORD 6629 (1959). Availableto qualified requesters from DDC, Cameron Station, Alexandria, Va.

12. Trueblood, K. N., Goldish, E. and Donohue, J., Acta Cryst.,14, 1009(1961).

21

UNCLASSIFIED

DOCUMENT CONTROL DATA . R & D

SG.,A N A t , r[iraie iir) , I 'f VOa R 5 Cik) W tTY C LA-,IFIrfAtI rO.

U. S. Naval Ordnance Laboratory UNCLASSIFIEDWhite OaK, Silver Spring, Maryland .

P0,~

Vapor Pressures and Heats of Sublimation of Some High MeltingOrganic Explosives

4 [)[.scrll ý Ii VF NO T Ts , , tCit 4 prt arid inchl-i delIties)

qAL ti 08 O 17 itr fr t1,r. - tItt v t initia . , I -tutu;

Rosen, Jerome M. and Dickinson, Charles

PI R T rt ~AT E t. TO TAt NO OFt P0- AGF S ~16 April 1969 33 12

6. C tN TiAC ORt TiA *O i i .AO 5REPi-'ii T -t~ii -51

h ,FCoRO ORD-332-001/092-1/UFl•7- NOLTR 69-67354-301

MAT 03L OOO/Z..RC (I -.l I'I . .... .. o .i ....) Aii..i, .. . t i. .. ,. , .....

This document has been approved for public release and sale, itsdistribution is unlimited.

Naval Material CommandNaval Ordnance Systems Comrand

'The Langmuir method was used to determine the vapor pressures and heatsof sublimation of several high melting organic explosives. The tempera-tures at which each compound would have a vapor pressure of 10-7 torrare: (1) cyclotrimethylenetrinitramine (RDX), 48.o 0 c; (2) cyclotetra-methylenetetranitramine beta polymorph, (8-HMX), -21.9 C; (3) 2,4,6-trinitroaniline (TNA), 44.O°C; (4)l,3-diamino-2,4,6-trinitrobenzene,form I polymorph, (DAýB-I). 79.8°C; (5) 1,3,5-triamino-2,4,6-trinitro-benzene (TATB), 1-,i.4 C; (6) ';,21,4,,6,6t.hexanitrostilbene (HNS),170.6 0 C. Wtthirn a series those compounds which form the strongestintermoleculr hydrogen-bonds have the highest heats of sublimation.

AN

DD Fo. 1473 UNCLASSIFIED

UNCLASSIFIEDSecurity Classlu icetion

KE WRO fLINK Al LINK 8 LIN, C

14K Y 1ORISkLa - I ROLE P OLE_ IVT

Vapor PressureHigh MeltingOrganic ExplosivesHeat of SublimationRDXe-HMXTNADATB-ITATBHNS

DD h1o 1473B UNCLA•.sIcm E _____

PPA, } ? Secuitiy Caa6 ricasia'-