Crystallographic Data. 32. RDX (Cyclotrimethylenetrinitramine)

---

ANALYTICAL CHEMISTRY

934

room),, which is a standardization of the meter circuit. The meter IR Bet for the range from 0 to 100 mw. per square foot, the shutter i8 oDened.. and the lieht intensitv is adiusted bv means

known amounts of mercury in 250 ml. of water through the complete procedure. ACKNOWLEDGMENT

Thanks are here extended to Warren Ksmm and Howard MoKinney for assembling the apparatus and preparing the curves. pa.d--!iaidinf? I r k pleated by mean8 of'tbe movablefurnace, and

LITERATURE ClTED

(1) Ballard and Thornton, IND.ENO.C~EM., ANAL. ED., 13, 893

(1941).

The

CI'PVCS

given in. Piqures 2 and 3 were obtained by taking

RECEIVED January 16, 1950. Communication 1310 from the Kodsk Research Laboratories.

32* B.BX (Cyc1otsi.methylenetrinitrarnine) Contr5,loixtsrl 5 y ?VA?,TE!$ C MCCROWF, Armour Research Foundation, Illinois Institute of Technology, Chicago 16, 111.

CH,

/\

d 2 25 2 20 2 12 2 08

I/Il Very weak Very weak 0 04 0 04

-

I/Ii 0.07

d 2.01 1.97 1.92 1.86

0.09 Very weak 0.07

.

Structural Fornula for RDX

"-

DX is the symbol for the high explosive cyclotrimethylenetrinitrsmine, which was developed for u ~ eduring World

War 11. It can be crystdliaed from a variety of solvents includiiig benzene, nitrometlmne, acetone, acetic acid, and nitric acid. It possesses one very unstable polymorphic form (II), which e m be isohted only in small quantities for a few seconds during fusion

studies on a xdcroscope slide. Becmsuse of ita physical instability, there is no possibility of.obtaining RDX I1 even in laboratory recrystallisations. EDX (1)

CRYSTAL MORPHOLOGY Crystal System. Orthorhombic. Form and Habit. RDX crystallizes in 8 wide variety of habits from needles (nitric acid) and plates (acetic acid) to massive (nitromethane acetone). Usually flattened on 001 showing theforms: (1101, (1201, 11011, {.Olll,md (1111. Axial Ratio. a:b:c = 0.881:1:0.813. Interfacid .Angles (Polar). 1104 110 = 32'50'; 1 2 0 A D O = 590; 101 A io1 = 850 30'; 011 n 011 = 780 20'. X-RAYDIFFRACTTOE DATA Cell Dim~ensions. a. = 11.61. A; h = 13.19 A,; e = 10.72 A. Fomula Wejshts per Cell. 3. Parmula Weight.. 222.13. i3nnsity. ?.?2 (pyoiomster); 1.81 (x-ray).

c

d

Figure 1. Photomicrographs of RDX

V O L U M E 22, NO. 7, J U L Y 1 9 5 0 a

955 The melt solidifies spontaneously and rapidly ~ i t hlarge gas bubbles showing a terracelike structure. Many of the larger areas of uniform orientation show an off-center arute birer interference figure with 2E = 113", 2 H = 67" I ' characteristic of R D X (11). Usually, honevv~, 1 pears showing centered acute bisectrix or off-cwitoi 'll)ric i figures with 2E = 92', 2V = 53", r > ii ( - j.

a

I

I

j

RDX (11)

Figure 2. Orthoof graphic Typical Projection Crystal

b - > + <

of RDX C

OPTICAL PROPERTIES

ACKNOWLEDGMEhT

Dispersion of RDX (20' C.) ( 4 ) Wave Length of Light, A. 6680 5893 5350 5020 4860 4710 4470

Refractive Indexes Alpha Beta Gamma 1.5906 1.5966 1.6031 1.6C84 1.6113 1.6145 1.6205

1.5725 1.5775 1.5827 1.5874 1.5895 1.5923 1,5970

An unstable polymorph of R D X can be obtained by recrysta!lization on a microscope slide from high boiling rolvents qucli 2 thymol, nitrobenzene, aroclor, and TN'I'. This modification has not been chsracterized because of the difficulty of obtaining and preserving well formed crystals. It should never he obtained on a laboratory scale; hence it should not interfere mitb analytical schemes based on the crystallography of RT)X (1) Some of the optical properties are given above iiiidi,i Fn'ioii Data.

1.5957 1.6015 1.6079 1.6130 1.6157 1.6187 1.6248

It is a pleasure to acknowledge the assistance of Irenc Corvin

Optic 4xia1 Angle 2E 2v 95" 15' 91'36' 87'33' 83" 53' 82' 1' 79'42' 75' 34'

55'21' 53'22' 51'8' 49"7' 48'4' 46'47' 44" 26'

Dispersion. r > b. qptic Axial Plane. 100. Sign of Double Refraction. Negative. Acute Bisectrix. c . = 1.592. Molecular Refraction ( R )(5893 A.; 25' C). i%F R (calcd.) = 43.7. R (obsd.) = 41.4. FUSIONDATA. R D X sublimes readily before melting with decomposition a t 205' C.; platelets lying on 001 are formed.

in determining and measuring the powder diffraction data, and of John Krc in the single crystal x-ray work. Most of the Kork reported above was carried out a t Cornell University under contract (B-123, OEMsr-193) with the Office of Scientific Research and Development (2). BIBLIOGRAPHY (1) Hultgrean, R., J . Chem. Phys., 4, 8 4 (1936) (2) Office of Scientific Research a n d Development, OSRD Eepf 694 (July 15, 1942), declassified Dec 5 , 1946. (3) Soldate, .4.hl., a n d Noyes, R. RI , ANAL CHEW,19, 442 ( I R 4 i ) (4) Terpstra, P., Z.Krast., 64, 150 (1926). CONTRIBUTIONE of crystallographic data for this section should he sent to Walter 6 . McCrone, dnalytical Section, Armour Research Foundation of Illinois Institute of Technology, Chicago 16 I11

Titration of Functional Groups with lithium Aluminum Hydride SIR: Since the publication of the article, "Electrometric Titration of Alcohols Using Lithium Aluminum Hydride" [Lintner, C. J., Schleif, R. H., and Higuchi, T.,ANAL.CHEM.,22, 534 (1950)], further studies in this laboratory have shown that molecular oxygen reacts rather rapidly with lithium aluminum hydride in solution. This is indicated in Table I, which shows the decrease in the effective hydride concentration in solution when dry oxygen is bubbled through the solution. This effect was not previously noticed in our analytical studies, as compensatory deterioration of the reagent probably took place in the blank determination. Furthermore, the hydrogen gas which is usual11 produced during the reaction period probably acted to a certain extent as a gas blanket. Table I.

Effect of Oxygen on Hydride Concentration Bubbled with Dry Nitrogen,

N

Open to Air, N

Bubbled with Dry Oxygen, N

Immediate titration" 0.538 0.499 0.420 Titration after 5 minutesb 0.539 0.427 0,082 Lithium aluminum hydride aolution pipetted in and titrated immediately with magnetic stirring. b Solution titrated after standing for 5 minutes under conditions indicated and with magnetic stirring.

Our analytical procedure has, therefore, been modified to reduce the effect of this oxidative side reaction. All analyses, including the blank, are now run under nitrogen. The hydride solutions are also stored under nitrogen.

TAKERE HIG~CHI School of Pharmacy University of Wisconsin

Madison, Wis.

Constant-Rate Feed Device SIR: In connection with the note on a constant-rate feed device [ANAL. CHEM.,22, 626 (1950)], I would like to call attention to an article on the same subject [Can. Chem. Process I d s . , 31, 153-4, 157 (1947)l. Basically the meters are the same. The one we have described contains two refinements which could readily have been incorporated into that of Lundsted and associates. Building the manometer into the flowmeter itself as we have done is only a minor improvemenf, However, we believe that use of the Mariotte principle by extending the air-inlet tube to the bottom of the liquid reservoir results in a major improvement in control.