Some &ects of variations in polymer cornpition are here shown by teat data on MllcsnisateS of Buna N types ranging from 26 to 46 per cent acrylonitrile. A formula containing 50 parts Bemireinforcing black, with normal sulfur and aceelera. tor concentrations but no added softener, was used in this work. In Figure 8 (left) the volume change in gasoline of these comwnnds. containinE various nitrile content wlvmers. is shown as a function f the aromaticity of t h i &&on medium. It is evident that an incraase in nitrile content of the polymer is reflected in a corresponding improvement in resistance to swelling in each gasoline. S l o p of the curves of volume incraase us. aromaticity for the several compounds are approximately aimilar. Improvement in resistance to solvent swelling with increme3 nitrile content is not obtained, however, without corresponding impairment of flexibility d low temperaturas. In Figure 8 (canter) curves of converted Shore hardness ue. temperature for the same compound series show th&he high nitrile-content polymera yield vulcanisatea which rapidly approach brittlenass with decreased a Shorehardnass the 48per cent nitrile content compound of 90 at 0' C., whereae the compound containing 26 nitrile polymer attaina the same hardness at -25'
Lawrence A. Wood, Norman Bekkadahl, and Frank L. Roth Vational Bureau of Standards, Washinfion, 0.C.
asoline is shown in Fi
ernod has been developed for preparing specimens of synthetic rubber in a suitable+ precise measurements of the ity. The rubber IS outgassed in a vacuum chamber and, while still under vacuum, is molded into a *et about 1/16 inch thick. Specimens weighinrabout 1 gram each are cut from this *eet and are employed for the measurement of the density by the method of hydrostatic weighings. The values obtained with different specimens from the same sample rarely differ from one another by more than 0.05 per cent. Values dre given for the mean densities obtained from six specimens made from single samples of each of eighteen of the most important commercial varieties of synthetic rubber now made in the United States.
of hardnass at -20' C. us. &maticity deceases with increased nitrile content. Then, even in contact with high aromatic content gasoline, the low-temperat&exibity of very high nitrile Buna N types is relatively poor. Thus,these data indicate that, for the Buns. N types,)Che properties of &stance to solvent swelling and resistance to hardening with decreased temperature are inherently interdependent functions of the acrylonitrile content of the polymer. This is most clearly shown here by the c w e s of volume change and freeze resistance us. polymer composition (Figure 9). Acknowledsment
The writers are indebted to many of their associates for assistance in the preparation of thia paper. Literature Cited Bryan. V., and Habgmd. B. J. (to Imperial Chemiod Induatries). Brit. Patant 498.302 (Jan. 2. 1939). Carman, F. A., Powers, P. 0.. and Robinaon, H. A., h. ENQ. Cmu.. 32, 1-72 (1940). Du Pont de Nemoura, E. I.. & G., ha.. Rubber Chem. Div., Supplement to Rewrt 2W (Nov., 1938). mebasttai, c. A, ~ndiaR&& w m .102, NO. 6. a3-s (1940). Ibid., 102, No. 6, 37-9 (1940). Ibid., 103. No. 1. 46-8 (1940). Koch. Albert. IND. ENa. CHI&, 32,464-7 (1940). Kwh. E. A., Kautwhuk, 16, No, 12, 161-0 (1940). Lightborn, I. E., Bubbw Age (N. Y.),47,19-21(1940). MoCortney. W. J., and Hendriak. J. V.. IND. E m . Cmx., 33. 679-81 (1941).
S.A. E. Journd, 50, No. 0.20-2 (1942). 8elher. M. L., Winspear, G. G.. and Kemp, A. R.. IND. ENa. CHIu.. 54, 16740 (1942). Stsom Eiatributora Ino.. "Perbunan. Gmooundinn - and Prw-
eming-. 1942. (14) Williams. Ira. snd Bmith, C. C. (to du Pont de Nemoura & (3.). U.8. Patant 2,064,680 (Dee. 16.1930). (16) Wmd, L. A.. Nstl. Bur. Standards, Circ. 0127 (1940). (16) Yersley, F. L..and F-. D. F.. h. &a. c-.. 34. a a w (1842).
OR many purposes the density of a material need be known with an accuracy of only a few per cent. For many materiale the density of one sample may mer from that of another by this amount, and no useful purpoee may be served by making more precise measuremente on individual samples. In natural rubber the densities of Merent samples have been shown, in a compilation (2) of twenty-one values, to lie with two exceptions between 0.905 and 0.919 gram per cc. at 25" C. The variations probably represent real merences in the samples and not accidental errom of observation. Since one can seldom know exactly the origin and subsequent treatment of a sample of natural rubber, there is little d u e in i n c r d g the precision of mwurement. Synthetic rubbers, on the other band, can be regarded as ususlly produced under conditions which are much better controlled and known. It is logical, then, to measure the density with greater precision and to hope to be able to ascribe significance to its variations from one sample to another. The present work is concerned with the development of a
INDUSTRIAL A N D ENGINEERING CHEMISTRY
method for preparing specimens of syntheticrubber in a form suitable for Dreciae mwurementa of the density, the actual measurement by the method of hydrostatic weiddngs, and a premntation of the reaulta for most of the varieties of synthetic rubber now of commercial importance in the United States. Preparation of Specimens
Samples of synthetic rubber, as d v e d from the manufacturer, were never found to be in a form suitable for precise measuremente of density. Even where there appeared tobe no entrapped air and the surfaces ~ e e m e to d be relatively smooth, molding produced specimens yielding higher and much more consistent values. Since a number of precautions are neceasarJ. order to obtain good s p e c k , the method of preparation will be d e a c r i t d in some detail. The ease of preparation of specimens was considerably dected by their rbedogical properties, and there was a wide m e r enm between the different types of synthetic rubber in this r e s p e c t . I n many ceaea not all the precautionsdeacribed were necmxy. The surface of the material was first examined for the presence of talc or other DOwderS which might have been dusted-on it. Vigorousscrubbing with a brush under running water usually removed the powders. This was followed by drying in a vacuum desiccator. About 7 cc. of the material were then sheeted out by several passesbetween warm mill rob. The sheet, usually between 0.010 and 0.015 inch thick, was folded over i t d several times a n d placed in a mold consisting merely of a sheet of metal with a single rectangularopening. Thespecimenand mold were sandwiched between aluminum sheeta and placed in E the vacuum chamber shown in Figure 1. ' n e cnamwr was of steel and inc
VoL 34, No. 11
but rolled as the piston moved up or down. The gasket was about 100mm. in diameter and about 4 mm. in seation diame ter. Four springs at the corners of the plate at the top of the chamber exerted an upward force on the piston. They were so choaen that, when the chamber was evmuated, a t mospheric preasure on the piston moved it downward about half ita full stroke, compressing the springs. The chamber was then placed between the platens of a vulcanizii p r w , and the platens were moved togeuler SufEciently to make contact with the tap of the vacuum chamber and to compress the springs a little more, but not enough to move the piston far enough to cause appreciable flow of the material. In this manner the material was heated for 1 or 2 hours at about 130' C. under vacuum. Next the p l a h of the hydraulic press werebroughtsomewhat clom tcgether to causesomeflow of the material. After several ntages of increase of hydraulic p m , the full force of 50 tons was b l l y applied to the vacuum chamber. After 1 or 2 hours to allow for further flow, the steam was turned &, and the press dlo*ed to cool with the hvdrsulic pressure still ipplied. Finally air was admitted to the chamber and the specimen re-
- . .
same material wi about 3.5-m~i. clearance between cbamber andpiston. The piston was sealed by a toroidal rubber gasket which did not slide on the pistan or side walk,
z! Figure 1. Vacuum Chamber Section M'is shown at the lower lek. The cross section of the rubber arket is marked in black. The Iates at the to and bottom ofthe cfambcr are 180 mm. square. f i e diameter opthe piston is
moved. Somematerials could be separated only with difficulty from t h e
aluminum sheeta above and below the specimen. In some c&898 mold lubricante, such as Carbowax, when used in small amounts were found to eliminate thk di5oulty. Carbowax is water soluble and was readily removed from the snrface of the specimens by washing. Chemigum I and some other materials were separated from the aluminum most easily when cooled t o a b o u t
The dimensions of the sheets made in this manner were about 60 X 60 X 1.6 mm. (2.4 X 2.4 X '/~ainch). E d sheet was cut into three specimens with soissors. In cutting, care was taken to discard portions of the sheet with surface irregularities and to cut with a single stroke so as to give a smooth edge. The specimens were d c i e n t l y thin so that with most of the materials inhomogeneities, if present in the interior, could w i l y be sean when the specimen WBB held
INDUSTRIAL A N D ENGINEERING CHEMISTRY
up to the light. Specimens with inhomogeneities were discsrded. Measurement by Hydrort.tic Weighing I Neoprene CG DU Pont Co. 1.2807 I Nooprane E Du Pont Co. 1.2384 The measurement of density by the method of hydroahtic I Du Pont Ca. Neoprene FR 1.1406 I Du Pont Co. Noopren. GN weighing required the use of a sinker, since many of the mate1.22w IIA fiatone Tire & Rubber Co. B u s8b 0.9291 rials were lighter than water. The sinker employed in the IIA . BUM8. fieatone Tire & Rob& Co. 0.98E4 IIA BUM 8 U. 8. Bubbsr Co. 0.9869 present work was a tin dieJc about 20 mm. in diameter and IIA BUM 8 Bt.ndud Oil D w e b ment Co. 0.98W G w d y ~ rTire & Ru%br.Co. IIA about 4 mm. thick, to which the specimen could be conven0.9391 IU B. F. Gwdriob Co. 0,0386 iently attached. The plane of the specimen was kept appmxiIIA B. F. Goodrioh Co. 0.8808 IIB Gmdyur Tire & Bubbsr Co. 1.0186 mately vartical to diminish the resiatsnce to the motion IIB B. F. G d o b Co. HsOR 0.9992 IIB Perbun Standud Oil Dedopment Co. through the water during weighing. A 6mil platinum wire 0.96E4 IIB Tbbkd RD Thiokol Cmp. 1.0604 suspended the sinker from the hook above the pan of the I11 Thiokol Corp. Tbbkol A I.69ea I11 Thiokol PA Thiolrol Cor . 1.3298 balance. The water waa contained in a besker mpported IV Butyl 51.46 8t.ndrrd d D w e b p m e n t Co. 0.9176 on a small wooden bridge over the pan of the balance. Before eaoh set of measurements the platinum wire was w d e d with water containing talc or soap so that it could be wetted readily and the deeds of surface tension minimised. The specimen was next pIaced in the water and then removed. Usually air bubbles were observed din& to the specimen, particularly on the edgea. Them were rubbed ofi with the fingers, and the specimen was r e p l a d in the water. The operation was repeated, if necewy, until no air bubbles could be seen. weighin@ were &ed out with an ordinary analytical balanceto 0.1 or 0.2 mg. The quantities o b e d and the method of computation rue illustrated by the typical data and calculations shown 1 in Table I. The temperature seldom d8ered from 26’ C. by more than 2’ or 3’. Consequently it was seldom necesssry to have accurate values for the rate of change of density of th% sample with temperature. In some eases the rate waa determined frommeasurements of density at twoMerent temperatures, and in 0th- it was obtained from independent mensurementa of the expansivity by dilatometers. In few casesY waa the value significantly Merent from 620 X 10-6 gram were made within a half hour after heating. C r y d l h om.” (” C.)-l, the value reported (1,s) for natural rubber. &n deeds do not -dent with any of the other varieties of s theti bber listed in Table II. However, it was found at a n of the synthetic rubbers showed &tic recovery d e c k which made it preferable to take the measTable 1. Typiul Dab and GICUI~UOIII urements soon after molding. This elaatic “memory” led to Neoprene ON BUM 8 the development of roughened surfaces and sometimee to 1.1673 2.0677 11.1370 11,1866 vacuoles, which brought about a decresse in the apparent 11.0497 11.6266 density after a few days. The results under them condi0.3W -0.00-54 1.2431 1.8891 tions were always much leas consistent than those obtained 0.9810 1.2828 24.6 26.6 from observations made immediately after molding. The D a u i t k . zrnmdc0.d n k t y for molding under vacuum did not seem to be so 8. Wa* at temp. Q 0.0971 0.0988 I . 8.mple at tamp. Q 1.2293 0.9279 grest for those materide in which the “memory” effects were J . &mole at 26O C. 1.2190 0.92~0 small. It should be noted that, although the values given represent the means of obsenrstiom on a number cd specimens, the specimens were made from only one sample of eacb variety. Consequently this work furnishes no information regarding the variation in density from Merent batches or runs of the same manufacturer. The method of meaeureRrulb ment described here should be sufficiently precise to h d The results of measurements of the densities of eighteen application for this purpose and might be used as a control difierent varieties of synthetic rubber are given in Table II. method. With the exception of the sample of Bum S described in foob Acknowledgment note’, they were all of commercial origin and rue considered Grateful acknowledgment is made of the cooperation of typical of -&day commercial production in the United the companiee listed in Table I1 in furnishins samples of Statm. Most of them were received during the spring and synthetic rubber on which the observations were made. E& summer of 1942. The values shown are, in moat caw., the p i a l thsnke are due to J. N. Street for the preparation and means of the measurements on six d8erent specimens, three d y s i a of the polymer described in footnote”to Table II. cut from each of two moldimp. Additional specimens were molded when any of the individual reaulta mered from the Literature Cited mean by more than about 0.05 per cent. In a few eases only four specimens were used, and in other instancee more than six were made. Unpublished observations in this laboratory have shown