374
T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERIhTG C H E M I S T R Y .
bisulfide and separated from this solution with gasoline. This treatment has the double effect of removing the free bromine more exhaustively and converting the precipitate into a stable crystallized form. The crystals are oxidized with nitric acid in the original vessel in which they were precipitated. A sufficient, measured amount of standard silver nitrate solution is added and the mixture gently heated. The nitric acid has t o be replaced from time to time, and the manipulation continued until the bromine is all combined as silver bromide. The excess of silver nitrate is titrated back b y Volhard’s well known method with ammonium thiocyanate. The formula C,,H,,Br, forms the basis for the calcu1a:ion of pure rubber, four bromines corresponding t o one caoutchouc, C,,H,,. I n applying the method i t is preferable t o extract the sample first with acetone and eventually with alcoholic potash t o remove completely the resins which might prove a disturbing factor b y forming insoluble bromine compounds. Spence has only recently demonstrated b y his investigations on rubberproteins and their behavior towards bromine that the amount of the latter added to the tetrabromide in Budde’s method is too small to materially influence the final results, provided the tetrabromide is calculated from the amount of bromine contained in the precipitate. The same author observed a regular loss of bromine when oxidizing the *tetrabromide with nitric acid as proposed b y Budde, a fact which was confirmed b y Hinrichsen and Kindscher. They therefore suggested t o replace the oxidizing and titrating process b y a combustion in a current of oxygen or fusion with sodium-potassium-carbonate and potassium-nitrate mixture. The fusion method has been improved and more fully described b y Spence who obtained satisfactory results. Budde’s method, therefore, gives in the case of raw rubber, even a t its present stage a fair idea as t o the amount of rubber contained in a given sample. A more difficult problem faces the chemist in the case of vulcanized goods. The question, “What reactions take place during the bromination?” %ecomes more complicated and the composition of the products is a matter of great uncertainty ; the method is, therefore, practically valueless for the estimation of pure rubber in cured goods. Attempts to solve this problem were made b y two different investigators. Axelrod precipitated the tetrabromide from a solution in kerosene with Budde’s bromination mixture. The precipitate was washed in the same way as b y Budde, and dried a t from 50 t o 60’ C.; after weighing, it was ignited and the ash deducted. By analyzing a number of compounds of known composition he established a factor for the calculation of pure rubber from the bromo-sulfo compound. A moment’s thought will a t once reveal t h a t such a determination is hardly applicable for general use. All sulfur of vulcanization is included in the bromine precipitate, and, as the coefficient of vulcanization is subject t o considerable variation, it follows necessarily that the composition of the bromo-sulfur compound of rubber will be just as variable. We ha;e, considering the present lack of
May, 1912
a proper theory of vulcanization, no means t o judge what reactions take place and what compounds result when cured rubber is brominated. If the sulfur of vulcanization is combined chemically with the rubber molecule, there is, of course, no room left for a n addition of bromine t o the double linkages which are saturated with sulfur. Huebener took this question into consideration, and, by adopting the WeberDitmar theory of vulcanization, concluded the reaction product of bromine and vulcanized rubber consisted of three distinct compounds: tetrabromide, dibromomonosulfide. and disulfide. For purposes of calculation it is necessary t o consider only tetrabromide and disulfide because two molecules of dibromomonosulfide are equal to one molecule of tetrabromide and one molecule of disulfide. Therefore, not only bromine but also sulfur of vulcanization has t o be determined t o find the equivalents for pure rubber, He proposed a t the same time a somewhat simplified method, originally designed for hard rubber, which has the advantage of eliminating the often very troublesome and long proposition of dissolving the rubber. The sample is finely rasped and brominated with elementary bromine under water. The results obtained in our laboratory have been invariably high, as it seems practically impossible to free the precipitate from the excess of bromine b y a simple washing with hot water. Of course I have t o admit t h a t we have not made an exhaustive study of the subject and therefore are not ready to condemn the method as valueless. Even if all methods, as described before, have failed to be satisfactory to everybody and in every instance, their development marks a decided progress in rubber chemistry. It can be justly hoped t h a t a t a not too far distant time chemists will be able t o submit such methods of analysis as will be acceptable t o both the merchant and the manufacturer. The importance of this problem was recognized a t the International Rubber Exhibition at London when for the first time the International Testing Committee met in full session t o discuss the questions which should be settled by this body.
THE VOLATILE ACIDITY OF GUM TRAGACANTH COMPARED WITH THAT OF INDIAN GUM. BY W. 0 . EMERY. Received March 13, 1912. N E E D O F T H E INVESTIGATION.
The primary object of this investigation was to devise additional methods for the detection of Indian gum when substituted for or in admixture with tragacanth. A study of the literature early led t o the conviction t h a t among the degradation products of these gums there must be one susceptible of quantitative isolation and sufficiently characteristic t o serve as an indicator of the purity, and therefore of the quantity, of parent substance involved in its hydrolysis. Such a n indicator was believed t o be represented b y acetic acid, already isolated from one Indian
May, 1 9 1 2
T H E J O U R N A L OF I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y .
gum and reported as being present in another. The hope was entertained t h a t tragacanth, under like treatment, might fail t o yield this acid, b u t as will presently appear, experiment proved otherwise. As defined b y the United States Pharmacopoeia, tragacanth is the “gummy exudation from Astragalus gurnmifer Labillardiere, or from other species of Astragalus, family Leguminosae, appearing in ribbonshaped bands varying in size and from I to 3 mm. thick, or in irregular pieces of the same, long and linear, straight o r spirally twisted ; externally whitish, marked .by more or less pronounced longitudinal or eccentric lines or ridges * translucent, fracture, short, tough, rendered more easily pulverizable b y a heat of 5 o ° C . (122’ F.).” While it is possible t h a t the products known t o the trade as “Indian gum” and employed so extensively in this country may vary as t o their origin, it appears reasonably certain t h a t two gums a t least are justly characterized b y this term, namely those of Sterculia w e n s a n d Cochlospermum, gassypium, both of which find local application as substitutes for tragacanth. It is equally certain t h a t none of the so-called Indian gum has its origin in any species of Astragalus. Most of the Indian gum reaching this country bears little or no resemblance t o tragacanth. A careful comparison of the commodity with authentic samples obtained direct, both from London and from India, clearly indicates t h a t i t corresponds in all essential points to the gum of Sterculia urens. It occurs in irregular, striated, sometimes twisted, translucent, or transparent lumps, never in ribbon-shaped bands or leaves. I n view of such physical characteristics, successful adulteration or substitution of whole gum tragacanth with Indian gum is no easy matter, and yet the latter product is occasionally offered for sale in the bazaars of India as true tragacanth.’ It is, however, the powdered form of tragacanth t h a t presents t o the sophisticator more alluring possibilities. Owing t o its extended use in the arts as well as in medicine, coupled with a relatively high price for the better grades, adulteration with the cheaper Indian gum has within the past few years been of frequent occurrence; hence any method looking t o the detection and estimation of such adulterant must be welcome t o all desirous of obtaining pure powdered tragacanth.
375
when subjected to hydrolysis. None of the investigators of tragacanth, however, has, to the writer’s knowledge, ever considered, the possibility of acetic or other volatile acid constituting one of the hydrolytic products of this gum, due perhaps t o the fact t h a t i t is odorless and without acid reaction in aqueous suspension, conditions not obtaining in the case of the two Tndian gums herein considered. The remarkable property possessed b y the gums of Sterculia w e n s and Cochlosperlnum gossypiuvu: of developing an acetous odor when exposed to moist air has been commented on b y various authors; in fact, an examination b y Robinson1 of the hydrolytic products obtained with the gum of Cochlospermuw gossypizcnz developed the fact that an amount of acetic acid was formed equivalent to 14.4per cent. of the original gum. When gum tragacanth is heated with an aqueous mineral acid, as phosphoric or sulphuric acids, and the products of such action subjected to steam distillation, a n acid distillate is obtained from which acetic acid can readily be isolated in the form of its silver salt. The procedure employed is as follows: Treat 2 0 grams of whole gum first in the cold with 2 0 0 cc. of distilled water and I O cc. of sirupy phosphoric acid until completely swollen, then subject for several hours to the full heat of the steam bath, whereby the mass gradually becomes partially liquefied, then distil the product with steam and evaporate the distillate t o dryness in the presence of barium carbonate. Treat the residue with a little hot water, filter, and distil the filtrate, amountingto abuut 30 cc., with steani after the addition of 5 cc. of’sirupy phosphoric acid. On treating the distillate with silver oxid and filtering, characteristic plate-like needles were obtained, which on ignition proved t o be silver acetate, as is evidenced b y the following analytical data: 0.2048 gram of substance gave 0.132I gram of silver; calculation for C,H,O,Ag gave 6 4 . 6 5 per cent. of silver, 6 4 . 5 0 per cent. being found. There can, therefore, be no doubt that acetic acid constitutes one of the degradation products of tragacanth when heated with mineral acids. E S T I M A T I O N O F ;ZCETIC .4CID
Quantitatively, the acetic acid or, rather, “ volatile acidity,”a is estimated as follows: Treat I gram of the whole or powdered sample in a 700 cc. round-bottomed flask, provided with a long C H E M I C A L PROPERTIES O F G U M T R . 4 G A C I N T H . According t o researches carried on b y O’Sullivan,* neck, for several hours in the cold with I O O cc. of distilled water and 5 cc. of sirupy phosphoric acid until tragacanth consists of starch granules, cellulose-that portion insoluble in boiling water, cold dilute acids, the gum is completely swollen. Roil gent,ly two and alkalies-likewise water-soluble gun1 yielding a hours in connection with a reflux condenser, whereby series of gum acids of the nature of gedtlic acid, bas- a nearly clear, colorless solution is effected. A very sorin, bassoric acid, nitrogenous and mineral matter. small amount of cellulose substance will remain unAS the result of a n investigation of tragacanth from dissolved. Now subject the hydrolyzed product to various sources, Hilger and Dreyfus,3 among other slow distillation in a vigorous current of steam until conclusions, found t h a t this product differs not only the distillate amounts to 600 cc. and the acid resias regards chemical constitution, but also with re- due t o about 20 cc. This should not be driven too 1 J . Chen. SOC., Trans., 89, 1496 (1906). spect to the proportion of degradation products formed Zornig. Arzneidrogen, 1909, p. 654. a J . Chem. SOC.Trans., 79, 1164 (1901); Proc., 1901. p. 156. Ber. d. chem. Ges., 33, 1190 (1900).
*
*
The term “volatile acidity,” as used herein, is the number of cubic centimeters of tenth-normal potassium hydroxid required t o neutralize the volatile acid or acids obtained, hy subjecting the products to the action of aqueous phosphoric acid on 1 gram of gun1 to distillation with steam.
3 76
T H E J O U R N A L OF I.VDUSTRIAL A N D E.YGIh7EERI-I-G C H E X I S T R I ' .
far, however, otherwise there may be danger of scorching the non-volatile, organic degradation products, with consequent possible contamination of the distillate. I t has been found t h a t a spray trap of the form shown in Fig. I , if used in connection with the flask containing the hydrolyzed gum, is effective in preventing traces of phosphoric acid being carried over into the distillate. Titrate with tenth-normal potassium hydroxid in connection with I O drops of phenolphthalein solution, finally boiling the liquid under examination until a faint pink color persists. Run a control on same amount of distillate obtained b y a parallel operation, with onlission of gum, but using like quantities of other ingredients and observing the same conditions as in the test.
May, 1912
(cc. tenth-normci potassium hydroxid) : 3 . 8 , 3 . 5 , 3 . 5 , 3 . 5 , 3 . 6 , 4 1 , 3 . 8 , 3 . 4 , -1.9, 3 . 9 , 3 . 5 , 3 . 7 , 3 . 8 , 3 . 9 , 3 . 6 , 3 . 8 , 3 . 6 , 3 . 8 , 3 . j , 3 . 6 , 3 . 8 . This gives an average volatile acidity of 3 . j cc., equivalent t o 2 . 2 0 per ccnt. of acetic acid in the original g u m .
As regards the acetic acid content of Indian gum (CocI~losperiiiz~.ilo,q o s s y p i u ~ z ) it has already been shown by Robinsonr to yield 1 4 . 4 per cent. of this acid. That of Sterculia urens, however, though early recorded b y Guibourtz as emitting an acctous odor, has never before, to the writer's knowledge, been examined either as to the identity or the quantity of the acetic acid present. Five grams of Indian gum (Sterculiu ztrens) were therefore treated in substantially the same manner as outlined for the isolation of the silver salt from the hydrolytic products of tragacanth. I t may not be amiss t o note in this connection t h a t while t,ragacanth yields a practically colorless solution when boiled with aqueous phosphoric acid, Indian gum, on the other hand, gives a pink or rose-colored solution, a sure indication in a n unknown sample that this product is present. The acid distillate obtained gave with silver oxid characteristic plate-like needles, which, on ignition, yielded values corresponding t o silver acetate: 0 . 2 0 1 0 gram of substance gave 0 . 1 3 0 0 gram of silver; calculation for C,H,O,Ag gave 6 4 . 6 5 per cent., 64.67 per cent. being found. Gram samples of whole and powdered Indian gum (Sterculia w e n s ) were examined for volatile acidity, exactly as in the case of tragacanth, with the following results: Results o n g samples of I n d i a n gum, lump (cc. tenthnormal potassium hydroxid) : 2 6 . 3 , 26. I , 26.6, 2 7 . 7 , 2 7 . 1 , 28.3, 2 5 . 9 , 26. I , 2 6 . 8 . These data indicate an average volatile acidity of 2 6 . 7 cc., corresponding t o 15.91 per cent. of acetic acid in the original gum. The last of the nine samples, yielding 2 6 . 8 cc., was an authentic one obtained direct from the Indian Government; this acidity corresponds t o 1 5 . 9 7 per cent. acetic acid. Results 01 14 samples of I n d i a n gum, powdered (cc. tenth-normal potassium hydroxid) : 2 6 . 5 , 2 6 . 3 , 26. I , 2 7 . 3 , 2 6 . 3 , 26.9. 2 j . 4 , 2 5 , 6 , 2 7 . 7 , 2 6 . 8 , 26.6, 2 5 . 6 , 2 7 . 6 , 26.0. The average volatile acidity in this case is 2 6 . 5 cc., corresponding t o 1 5 . 7 9 per cent. of acetic acid.
FIG.1.
CONCLUSIONS.
With a relatively large number of authentic gums (Turkey, Aleppo, and Persian), both whole and powdered, the following results were obtained: Results o n 35 samples of gum tragacanth, lump (cubic centimeters of tenth-normal potassium hydroxid): 3 . 9 , 3 . 5 , 4 . 2 , 4 . 0 , 3 . 4 , 4 . 0 , 3 . 2 , 3 . 5 , 4 . 0 , 3 . 9 , 3.4, 3.4, 3.6, 3.4, 3 . 5 , 4.2, 4.0, 3.6, 3 . 3 ) 3 . 6 ,
From the foregoing data i t appears t h a t the volatile acidity or amount of volatile acid developed b y gram samples of tragacanth, on the one hand, and Indian gum (Sterculia u r e n s ) , on the other, is fairly constant, sufficiently so, indeed, t o serve, in conjunction with other well-known tests, as a very reliable criterion for estimating the purity or quantity of either alone or in admixture. The volatile acidity of Indian gum (Sterculia urens) as compared with that of tragacanth is nearly 7 . 5 times as great.
3.8, 3.4, 3 . 3 , 3.6, 3.4, 3.1, 3.3, 3.5, 3.4, 3.7, 3 . 4 , 3 . 7 , 3 . 7 , 3 . 2 , 3 . 4 . These figures represent an average volatile acidity of 3 . 6 cc. Assuming this acidity
t o be due solely t o acetic acid, it would be equivalent to 2 . 1 5 per cent. of acetic acid in the original gum. Results o n -2 I samples 05 g u m tragacanth, powdered
BUREAU O F CHEMISTRY.
u. s. DEPARTMENT O F AGRICULTURE. 1 L O C . Cit. 1 Pharm.
J.,16, 5 7
(1855).
