Quantitative Determination of Natural Rubber Hydrocarbon by Refractive Index Measurements RACHEL J. FANNING AND NORh1.4K BEKKEDAHL National Bureau of Standards, Washington, D . C . Hubbers from different sources vary considerably in rubber hydrocarbon content, from 7 0 7 ~or less for some of the wild varieties to about 95% for a good grade of plantation rubber. The quality of a rubber is measiired to a great extent by the percentage of rubber hydrocarbon. A new procedure for the quantitative determination of this hydrocarbon has been developed, which involves the measurement of refractive index of a solution of a known weight of
acetone-extracted rubber in a known weight of 1bromonaphthalene. From the observed data and from other previously determined or known physical constants, such as the densities and refractive indices of the rubber hydrocarbon and the solvent, the percentage of the hydrocarbon in the original sample can be calculated. This new- method gives results as good as or better than other existing methods, and is simpler and less time-consuming to perform.
T
removed, the mixture is stirred, and the beakers are replaced in the oven. The hot mixture is stirred a t intervals of about 10 to 15 minutes for an hour or more until a homogeneous solution is obtained. It is stirred while cooling, and then weighed. The refractive indices of the contents of each of the beakers are measured by means of an Abbe-type refractometer. Techniques might be developed, however, for the use of other types of refractometers. These measurements are made on several portion8 of the solution, and unless the averages of the indexes of each portion are reproducible to at least 0.0001 the whole determination is repeated. The solvent, in order to be suitable for the determination, should not have changed i n refractive index by more than 0.0001. All measurements of density and refractive index must be converted to values a t 25" C. before they can be applied to the equation for calculating the results. Adjustnient of t,he density of the 1-bromonaphthalene is m:ide t)y applying the density coefficient ( l / p ) ( d p / d T ) , of -673 X l o + per degree centigrade. Coefficients for adjusting the refractive indices are -162 X 10-8 for I-brornonaphthalene, and -434 X 10-5 for the mixture of rubber and 1-bromonaphthalene in the strengths used in the procedure just described. The rubber hydrocarbon content, in percentage by weight, in a sample of natural rubber ie calculated from the equation:
HE rubber hydrocarbon in a sample of crude natur:tl rubber can be determined quantitatively by measuring the refractive index of a solution containing a known n-eight of acetone-extracted rubber in a known Lveight of solution. The procedure requires a knowledge of the values of the densities and refractive indices of both the rubber hydrocarbon and the solvent. -4suitable solvent should be able to dissolve the rubber in appreciable concentration, preferably up to 40% by volume. The value of refractive index must be very stable, and must be unaffected by heating required to produce solution. I t must also have a refractive index very different from that of the rubber hydrocarbon. For the present study I-bromonaphthalene was used as the solvent,. In :iddition to rubber hydrocarbon, crude natural rubber contains moisture, resins, proteins, ash, and other minor const,ituents The moisture can be removed by heating, and the resins can easily be removed by acetone extraction. However, it is more difficult to separate the proteins, ash, and other constituents from the rubber hydrocarbon. The method of analysis presented here does not require this Beparation, but involves dissolving the dry acetone-extracted rubber in a solvent. The assumption is made that all the material remaining in the acetone-extracted rubber, with the exception of the rubber hydrocarbon, is insoluble in the solvent and consequently has no influence on the refractive index of the solution. A second assumption is that the refractive index of ii solution of rubber hydrocarbon in a solvent is a linear function of the volume percentage of the rubber hydrocarbon in solution. These two :issumpt3ionsappear to have been justified by the results obtained from measurements made on a selected group of samples of natural rubber, including plantation rubbers, purified rubber, :ind a variety of wild rubbers.
in which iM is the weight of the niixture of extracted rubber and solvent in the beaker, E is the weight of extracted rubber in the mixture, D is the densit,y of the solvent in grams per nil. a t 25' C., R is the weight of the sample of rubber before extraction with acetone, ns is the refractive indes of the solvent a t 25" C., and TLM is the refractive indes of the nlisture of rubber and solvent a t 25" C. PHYSICAL CONSTANTS
PROCEDURE
The sample of rubber used for the analysis should give about 1.1 grams of rubber after extraction with acetone. It is either sheeted to about 0.5 mm. in thickness or cut into small pieces and weighed. It is then extracted with acetone according t o the method of the American Society for Testing hlaterials ( 1 ) . The acetone-extracted sample is dried in a vacuum oven for about 1 hour a t 100" C. and weighed again. A 30-ml. beaker containing a stirring rod is weighed; and the rubber sample, after having been cut into small pieces, is put into it. By mean$ of a pipet, or some suitable measuring device, about 3 ml. of 1bromonaphthalene are added to the beaker containing the sample. -4second beaker of similar size containing the same quantity of solvent, but no rubber, is prepared for the purpose of checking the constancy of refractive index of the solvent during the process. The index of refraction of the 1-bromonaphthalene in the beaker is measured, and the temperature is noted. The two beakers with their contents are covered with watch glasses and placed in an oven a t 140" C. to facilitate solution of the mixture. After about 30 minute? in the oven the beakers are
The analytical procedure requires the knoivledge of sonie of the physical constants of the rubber hydrocarbon and the solvent, 1bromonaphthalene. These constants for rubber have been surveyed by Wood ( I S ) , who selected from these values the ones which he thought to be the best. For the solvent, l-bromonaphthalene, reliable values were not found i n the literature, and therefore had to be determined. Rubber Hydrocarbon. The density of purified rubber hydiocarbon has been given by McPherson ( 9 , 1 9 ) as 0.9060 gram per nil. a t 25" C. Thisis somewhat Ion-er than the 0.911 value taken €or commercial raw rubber ( I S ) , but this is to be expected because of the absence of proteins in purified rubber. The densities of the proteins are probably greater than unity. The refractive index of highly purified rubber has been measured by JlcPherson and Cumniings ( I O , I S ) and found to be 1.5190 for the D line a t 25'C'. In the present investigation a sample of purified rubber was also prepared, and refractive indices made on this rubber by the method of Arnold, AIadorsky, and Wood (3)\yere found to laiige between 1.5158 arid 1.5190.
1653
1654
ANALYTICAL CHEMISTRY
These results agree very well with that given by McPherson and Cummings, and the value of 1.5190 is suggested for use in the present procedure, as the rubber prepared by the earlier investigators was the more highly purified. Wood and Tilton (14) in measurements on unpurified rubber found the rate of change of refractive index with temperature to be -370 X IOp6per degree centigrade. The method they used is capable of greater precision than that employed by the other investigators, and consequently their value for rate of change of index i s used. Solvent. The density of the solvent 1-bromonaphthalene was determined by the use of a picnometer. As several samples varied in density between 1.475 and 1.489 grams per ml. a t 25" C., it can easily be seen that the density must be determined for each new batch of solvent used. By the method of volume dilatometry (6) the expaneivity of the 1-bronionaphthalene was determined to be 673 X 10" per degree centigrade between l o o and 4 0 ° C This value was therefore used to determine the density a t 25' C when measurements u ere made a t other temperatures Different batches of 1-bromonaphthalene used in this investigation had refractive indice8 varying between 1.6547 and 16567 at 25' C. By measurement of the index of a given sample a t temperatures ranging from 10" to 60' C., a value for the temperature coefficient was found to be -462 X per degree centigrade. Corrections to 25" C. were therefore made using this figure
kRHC ;-
SOLVENT
too
0
RUBBER HYDROCARBON CONTENT,
VOLUME PERCENT
Figure 1. Schematic Diagram of Relation between Hefractive Index and Composition of Solutions Mixture. Refractive index measurements made on the rubber-solvent mixtures must be adjusted to 25" C. The values for the temperature Coefficient of refractive index as previously reported, -462 X 10-6 for 1-bromonaphthalene and -370 x for rubber hydrocarbon, yield a coefficient for a 30'% misture by volume of the rubber in the solvent equal to -434 X 10-6 per degree centigrade. DISCUSSION OF PROCEDURE
In the equation prenoudy given the factor ($1 - E ) / D represents the volume of the solvent in the mixture. The factor (ns - n ~ ) / ( n 1.5190) ~ represents the ratio of thevolumeof the rubber hydrocarbon in the misture to that of the solvent if linear relationship is apsumed, as it can easily be seen from Figure 1 that in the similar triangles the ratio of the difference of ordinates (refractive indices) is equal to the ratio of the difference of abscissas (volumes). The second factor, obtained from refractive index measurements, when multiplied by the first factor, therefore gives the volume of rubber hydrocarbon in the mixture. The product, when multiplied by the density of the rubber hydrocarbon, 0.9060 gram per ml., gives the weight of the rubber hydrocarbon. Thk weight, multiplied by 100, and divided by the weight R of the original sample, givea the per cent by weight of rubber hydrocarbon in the original eample.
Table I.
Results of Analyses of Various Types of Natural Rubber
Type of Natural Rubber
Rubber Hydrocarbon CAOb method method
RI" %
%
Acetone Extract
%
Total Accomnted InRI. ChOb soluble6 method method
%
Ribbed smoked sheet 94 1 93 2 8 ti 2 2 Blended smoked sheet 93.9 94.6 3 3 2.1 l7.S.F.C 3 0 93 9 93.9 L 4 93.3 Hmea benthamiana 94 7 2 8 2.1 81 .5 Heuea p a u e z f l o r n 80 7 16 7 2.0 Castilloa 84 9 85.2 11 6 2.4 7. 7. Landolphia 89 0 89.3 1.4 Funtumia 88.9 88 0 6.4 2.5 Clitandra 89 6 5.8 89 3 1.6 85 7 Ficus uogellz 85.3 8 6 3.7 Purified 1 97 5 0 98 8 Purified 2 98 9 ... 0 a Refractive index method as described in this paper. b Chromic acid oxidation method ( 7 ) . c Special rubber prepared by United States Rubber Co. d Sample contained 0.8% ash and 0.3% proteins. Sainple contained 0.47, ash and 0.3% proteins.
%
%
99.9
99 0
99 99 99 99 98 98 96 99 98 99 99
3 3 6 4
9 1
9 7 0 9 6
100.0 99.3 98 4 100 2 99 2 98.4 97.8 100 0 97 6 98 6
Care must be taken that the precision of measurement of the various quantities is adequate. The weights and the density should be determined to a t least one part in a thousand. The extracted rubber sample must be dried to constant weight, so as not to introduce an error in this weight. The presence of acetone, however, causes no error in the reading of the refractive index of the mixture, as it was found that the heating drives off all traces of acetone. The densit.y of the 1-bromonaphthalene solvent should be redetermined each time a new batch of the solvent is used, as it varies considerably from one bottle to another, even when furnished by the same manufacturer. Because of the subtractions involved in the second factor i n the equation, the refractive index measurements must be made as precisely as possible. A good general reference on refractometry, such as that of Bauer and Fajans (4),should be consulted for techniques of measurement. If five to ten measurenients are made on each of several portions of the sample, an average value for the refractive index can be obtained by means of the Abbetype refmctometer, for which the precision is better than 0.0001 (3). The fifth decimal place should hp retained for purposes of calculation. If the refractive indey value of the solvent has undergone a change by more than 0.0001 a? a result of heating in the 140" oven, the use of this solvent is questionable in view of the evidence of some permanent chemical change in it. When such a change is noted the batch of solvent should be purified or a more stable batch obtained. From a number of different aupplies of 1bromonaphthalene tested, it was found that in most rases the solvent was suitable for this work as received, but in a few inetances there was a change of refractive index so great that it was thought inadvisable t,o use the solvent without purification. The refractive indices of several port#ions of the mixture of rubber and solvent should be measured. If the average values of several portions do not agree with each other, there is indication that the solution is not uniform and that the stirring or the time of solution was not sufficient. The size of the rubber sample to be used for the analysis and the quantity of solvent for solution of the rubber hydrocarbon are dependent on several factors. Figure 2 presents a curve which shows the percentage error in the determination of rubber hydrocarbon caused by an error of 0.0001 in reading the refractive index of the solution. The greatest precision is obtainable for a mixture of equal quantities by volume of the rubber hydrocarbon and the solvent. Solutions of this concentration are, however, so viscous that it is difficult to stir sufficiently to produce a uniform mixture. Moreover, the great stickiness causes difficulty in separating the p r i s m of the refractometer after the readings have been taken. In one case the solution held the prisms together SO tightly that when sufficient force was used to separate them, one
V O L U M E 2 3 , N O . 11, N O V E M B E R 1951 of the prisms was broken loose from its position and had t o be reset. On the other hand, the error of the determination is increased if the concentration of the solution is too low. Concentrations a8 low as 30% by volume of rubber hydrocarbon do not appreciably affect the precision of the analysis, and the viscosity of the solution is sufficiently low to permit satisfactory manipulation both in the beaker and in the refractometer.
Figitre 2.
Error of Determination of Rubber Hydrocarbon
at Various Concentrations Caused by error of 0.0001 in refractive index of solution
I t w:i: f‘ound desirable to use about 3 ml. of l-bromonaphthafor each determination. This quantity does not need to be nw:~aurtui very precisely a t the t.ime of addition, but the use of a pi1)et was found to be very convenient. Some evaporation of the solvent takes place in the oven a t I4O0C., and therefore the ;tinount of solvent must later be determined by weighing. As the rate of evaporation of the solvent varies under different heating conditions, the quantity of rubber to be used to produce a mixture of‘ about 30% by volume will also vary. The weight of rubber situiple recommended for the analysis may therefore have to be ndjusted, depending on the particular conditions of test. Samples appreciably larger than that recommended here are extracted less efficiently by the acetone. If the samples are considerably smaller, either the volume of the mixture is too small for efficient stirring or the concentration of the solution is too low for good precision. Solution of the extracted rubber is greatly facilitated by cutting the sample into small pieces. I n a study made on the solution of rubber in 1-bromonaphthalene a t various temperatures it was found that, temperatures lower t,han 140 O C. increase greatly the time required for solution and also make stirring more difficult because of the greater viscosity. Solution a t temperatures considerably above 140”C. gave more rapid solution, but the results were abnormally high. It was thought that’ the protein molecules decompose a t these high temperatures, and then dissolve in the 1bromonaphthalene. I n order to obtain a uniform solution within a reasonable period of time without breakdown of the nonrubber conetituents, the directions of the procedure should be followed closely. leiit,
1655 rubber having a rubber hydrocarbon content of about 99% down to some wild rubbers containing only about 80%. The average deviation between values obtained by the two methods is about 0.6%. I n Eome cases the refractive index method shows the 1 igher results, and in other cases the reverse is true. There seems to bP a tendency, however, for the refractive index method to give slig’ tly lower results. In most instances, especially for the higher grade rubbers, the sum of the rubber hydrocarbon, the acetone extract. and the “insolubles” is close to 100%. This indicates t,’iat the values obtained for the rubber hydrocarbon content by the-e methods are close to that value obtained by the method of “differences” ( 5 ) . An attempt was made to purify rubber directly from preserved latex by a method based in part on that of Smith ( I l ) , in which the latex was first dialyzed, treated with sodium carbonate and trypsin to destroy the proteins, creamed several times after the addition of sodium hydrouide, and finally dialyzed again to remove the salts and alkali. The rubber was coagulated x i t h acetone containing phenyl P-naphthylamine(pheny1-2-naphthylamine), which was added as an antioxidant for the rubber hydrocarbon. The coagulated rubber,. which then contained some of the antioxIIC.AL ?-
SOCIETY, Washington, D. C., March 2, 1951. (9) McPherson, A. T., Bur. Standards J . Research, 8 , ‘751 (1932): Rubber Chem. and Technol., 5 , 523 (1932). (10) McPherson, A. T., and Cummings, A. D., J . Research .Vatl. B u r . Standards, 14, 553 (1935); Rubber Chem. and Technol., 8 , 421 (1935). (11) Smith, W.H., Saylor, C . P., aiid Wing, H. J., B u r . Standard8 J . Research, 10, 479 (1933); Rubber Chem. and Technol., 6, 361 (1933). (12) Willits, C. O., Swain, 11. L., and Ogg, C. L., IND. ENG.CHEM., AN.~L.ED., 18, 439 (1946); Rubber Chem. and Technol., 20, 320 (1947). (13) Wood, L. A , Proceedings of Rubber Technology Conference, London, England, May 1938, p. 933, Cambridge, England, \V, Heffer and Sons, Ltd.; Rubber Chem. and Technol., 12, 130 (1939). (14) Wood, L. A., and Tilton, L. W., Proceedings of Second Rubber Technology Conference, London, June 1948, p. 142; J . Research ;\‘atl. B u r . Standards, 43, 57 (1949); Rubber Chem. and Technol., 23, 661 (1950). RECEIVED April 13, 1451. Presented before the 58th Meeting of the DiviW’aahington, sion of Rubber Chemistry, AVERICAX CHEMICALSOCIETY, D . C., March 2 , 1951.
Measurement of Refractive Index of Elastomers AURELIA ARNOLD, IRVING MADORSKY’, AND LAWRENCE A. WOOD National Bureau of Standards, Washington, D. C .
F
It011 a careful study of the factors involved in tlie niearui’ement of t,he refractive index of elastomers, especially GR-S
synthetic rubbers, by means of a n -4bbe-type refractometer, a procedure for measurement has been developed which takes advantage of the inherent adhesiveness of most elastomers. This adhesiveness enables the polymer t o ‘ k e t ” the prism of the refractometer without t,he use of a contact liquid. The principle of the method is based on work done some years ago on natural rubber by RlcPhereon and Cummings (6). The detailed procedure has been developed in recent years by a number of different workers a t the Sational Bureau of Standards. Although the method, as given in this paper, has been described only in report form ( 2 , 7 , 8 ) , it has been used for several years as a basis for the determination of bound styrene in copolymers of butadiene and styrene ( 7 , l l ) ,and is currently employed for control purposes in the government-owned synthetic rubber plants. It has also been used by other investigators in deter1 Present address, Johns Hopkins Applied Physics Laboratory, Gilrer Spring, M d .
mirnng the per cent of rubbei hJdrocarbon in natural rubber (j), and for locating the second-order transition temperatures of natural and various synthetic rubbers (IO). APPARATUS
The conventional Abbe refractometer, graduated to the third and read t o the fourth decimal place, is used for the actual measurements. Water from a rrservolr a t room temperature is circulated through the housing of the measuring prism and through a cored brass block of about the same length and width a s the prism housing. The temperature of the circulating water is the same as that nf the specimen only when this temperature is near that of the ambient air. Consequently, it is preferable to allow the water temperature t o be approximately that of the room and t o reduce observations t o a standard temperature by calculation rather than t o maintain the water a t the standard temperature in a room that may be a t an appreciably different temprrature. The refractometer should be equipped with a thermometer graduatrd in units of 0.2” C. or less. If the temperature is not known to about this accuracy, errors will be introduced in the fourth decimal place of index. The standard glass test pieces furnished with refractometers
