REPORT OF THE SUB-COMMITTEE ON SHELLAC ANALYSIS. We

We have considered all the processes that have so far come to our 110- tice having for their object the determination of rosin in niixtures of shellac...
0 downloads 0 Views 423KB Size
[C O S T R I B G T I O S

F R O 3 1 THE COMNITTZX O N ~ S I F O R 3 f I T YI S

ASALYSIS.

TECHSICAL

IV. ]

REPORT O F T H E SUB-COMMITTEE ON SHELLAC ANALYSIS. Ey P A R K E RC. MCILHISEY. A . C. LANGXL-IR, ~ I A ~ I ~ I I L ITGCH, AS h l A X W A L L E R S T E I Y , X V G C S T U S H. GILL.

Receired M a y 31, 1907.

W e have considered all the processes that have so far come to our 110tice having for their object the determination of rosin in niixtures of shellac with rosin, and find that several of the methods depending upon the different powers of shellac and of rosin to unite with the halogens are capable of being used to determine the proportions of rosin and shellac i n t 11ei r mixtures. W e have not found any of the methods in which use is made of the other properties of these two substances, such as solubility, the acidity, or the Kottstorfer figure, sufficiently reliable for general use. As the absorption of halogen by both shellac and rosin is largely and perhaps entirely by substitution, and not as in the case of fats, by addition, and as the amount of halogen which so replaces hydrogen in the rosin is dependent upon the length of time that the reagent is allowed to act upon it, and also upon such other variables as the excess of the reagent, the temperature, and the relation of the volume of solution to the amount of substance treated, we have considered each of these halogen-absorption processes in the light of its adaptability to the maintenance of standard and uniforin arbitrary conditions which will permit results obtained a t one time being compared with those obtained a t any other time. T h e result of our work convinces us that the Wijs process' used as proposed by Dr. Langmuir, one of our number, possesses the desirable features of regularity and reliability in a greater degree than any other process which has come to our notice. It is possible by it to carry out a test of a saniple of shellac in a couple of hours, and as shellac is a rather expensive material and one very liable to sudden changes in price, it is almost essential that the standard method of testing it should be a rapid one. T h e Hub1 method is deficient in this respect, requiring usually 24 hours to carry out an analysis by it. T h e method which we would recommend as a standard one for the determination of rosin in shellac is substantially the one described in Dr. Langmuir's paper in the J. SOC.Chem. Ind., 24, 12. T h e solutions required are one of iodine monochloride containing 13 g. of iodine per liter, in glacial acetic acid that has a melting point of 14.7' to 15' C. and is free from reducing impurities, and another of sodium thiosulphate made by dissolving 24.83 g. of the pure salt in aliter of water. I n addition to these solutions there is required a quantity of acetic acid of the same strength 'Ber., 31, 7 5 0 , a n d J. SOC.Chetn. Incl., 17, 609.

I222

P A R K E R C. AICII.HXXEY,

ET. AI..

as that used for making the solutioii of iodine monochloride. Pure cliloro form and starch are also necessary. T h e preparation of the iodine niono chloride solutioii presents no great difficulty, but it must be doiie with car arid accuracy in order to obtain satisfactory results. There must he iii tli solution 110 sensible excess either of iodine or more particularly of chlo r i m , over that required to form the monochloride. This coiidition i nio.,t satisfactorily attained by dissolviiig i i i the whole of tlie acetic acic to be used the requisite quantity of iodine, using a gentle heat to assis the solution, if it is found necessary, setting aside a small portion of thi solutioii, while pure and dry chlorine is passed into the remainder unti the halogen coiitent of the whole solutioii is doubled. Ordiiiarily it wil be found that by passing the chlorine into the main part of tlie solutiot until tlie characteristic color of free iodine has just beeii discharged t h e n will be a slight excess of chlorine mhicli is corrected by the addition o the requisite amount of t h e uiichlorinated portioii until all free chlorinc has been destroyed. A slight excess of iodine does little or no harm, hu' excess of chlorine niiist be avoided. Introduce 0 . 2 0 0 g. of ground shellac into a 2 j 0 cc. dry bottle 0 : clear glass with a ground glass stopper, add 2 0 cc. of glacial acetic acic aiid warm the mixture gently until solution is coniplett A pure shellac is rather difficultly soluble. Solu. (except for the w a x ) . tion is quicker according to the proportion of rosin present ; I O cc. oi chloroform are added and the solution is cooled to 2 1 ' to 24'. The tetnperature should be held well within these limits during the test. 2 0 cc. of \Vijs solution are added from a pipette, having a rather small delivery aperture. T h e bottle is closed and stood in a dark place, and the time noted. It is convenient to keep the bottles during the test partly immersed in water which should be kept as nearly a s possible between 2 2 ' and 23". Pure shellac will scarcely alter the color of the Wijs solution. If in sinal1 amount, rosin will produce a slowly appearing red-brown color. I n large amount, rosin causes an immediate coloration, increasing in intensity as time passes. After one hour, I O cc. of I O per cent. potassium iodide solution are added. T h e solution is immediately titrated with t h e sodium thiosulphate solution ; 25 cc. or 30 cc. may be run in immediately, unless t h e shellac is very impure, and the remainder gradually with vigorous shaking. Just before the end, a little starch solution is added. T h e end point is sharp, as t h e reactiou products of shellac remain dissolved in the chloroform ; any color returning after a half minute or so is disregarded. -1blank determination should be run with 20 cc. of Wijs solution, 20 cc. of acetic acid, I O cc. of chloroform, and I O cc. of I O per cent. potassium iodide solution. T h e blank is necessary on account of the well

SHELLAC ANALYSIS

1223

known effect of temperature changes on tlie volume, and possible loss of strength of t h e Wijs solution. I n the case of grossly adulterated samples or in the testing of pure rosin, it is necessary to use, instead of t h e 0.2 g. of material, a smaller amount, say o . r j g. or even 0 . 1 g.. i n order that tlie excess of iodine n~onochloridemay not be too greatly reduced, as t h e excess of halogen is one of the factors in determining the amount of absorption. I t is safe to say that in case less than 2 5 cc. of thiosulphate solution are required, another test should be made, using a smaller amount of the shellac to be tested. I n weighing shellac. some difficulty is at times experienced on account of its electricnl properties ; i n very dry weather it may be found that the necessary handling to prepare it for weighing has electrified it and that it niay be necessary to leave it on the balance pan a t rest for a few niinutes before taking the final weight. I n any halogen absorption process it is essential that the substance to be examined should be tlioroughly dissolved in a reasonable volume of solvent that is cheap and easily obtainable. I n this case glacial acetic acid forms an admirable solvent ; the only other solvents which are a t ail comparable with it are the alcohols, which are too readily attacked by halogens to be used with perfect safety. The two defects which we have found in glacial acetic acid a s a solvent are, first, that it solidifies a t a temperature b u t little below t h a t of the ordinary rooiii. and, second, that the results obtained are to a certain e x tent dependent upon the strength of the acid used. On account of the first rnentioiied property it niay happen that different lots of glacial acetic acid or of Wijs solution may not be of uniform composition throughout the bottle 011 account of having been frozen and remelted; this difficulty can easily be obviated by remembering to shake the bottles of acid arid of Wijs solution before using t h e contents ; to a neglect of this precaution we are disposed to attribute irregular results which we have sometimes obtained during cold weather. T h e second difficulty is entirely obviated by using at all times acid of the same standard strength. Three prominent methods by which the strength of the acetic acid might be determined, namely, titration with alkali, the determination of specific gravity, and the determination of tbe melting point, we have investigated a s to their suitability for use in this connection. Of the three we greatly prefer t h e melting point determination, both on t h e score of accuracy and of speed. Titration we find least accurate of all, and such accurate determinations of specific gravity as would be necessary in making use of it as a criterion of purity would require an almost prohibitive ainount of time and attention. Furthermore, the ultimate standard of a melting point determination, an accurate ther-

rnoineter, is a much more satisfactory one and 011e more easily duplicated tliaii is the case iii tlie other two methods. lye have, tlierefore. in the description of the method, specified rather the ineltixg point of the acid to he used tliaii its strength as expressed by percentage. The iiiethod which we have used is a simp!e :ind rapid oue, Ijeing in fact the titre ' I method i n c0111111011 ase for deterniiniiig the inelting poiiit of fatty acids geiiernlly. A i l S-itich test tube is al)out half filled with the acid to be tested, t!ie coiiterits chilled rapidly to about 10' or 1 I ' and a delicate and accurate tliernioiiieter, grxiuated a t least to tenths of n degree, s:ispeiidecl i n the acid i i i tlie test tube, tlie latter being placet1 i n a wider test tube closed by A cork tlirougli wliicli the tebt t a b e coiitaiiiing the acid is inserted ; the \\.hole nrran,oemei;t is chilled ?)y being iiiiiiierseci iii ice lvater ; t h e acetic acid i n a few :iii!iutes begins to crysta:lize, and tlie temperature of the supercooled acid rises t o it. true nieltiug point. where i t reniairis statioiiary for sonie time. i i i tlic case of acetic acid of such s:rength as is used i n tlie process this hiniple procedrire gives the required itiformatioil i11 a few minutes ; \villi wenl;er acetic acid greatcr care is necessary to insure accuracy. 'The acid iii tlie test tube sliould iiot have begun to crystaiiize until some miiiiites after it has heen chilled in cold mater and mounted i n a larger tube : i f it does bc-gin to crystallize too soon, that is, before the cooled acid 113,- 113d sufikient t i n e t o become c;f uniform teinperature throughout, tlie acid sliould be well stirred with the thermonieter before tlie cr\.stal!iziiig process has gone very far. T h e melting point of acetic acid containi~igvarious percentages of water has been determined by Rudorff,' ns follows : I '

I k r ceiil. of r r n l

:icctic ncii:

3Ieltiiig p u i i i t .

.................. . . . . . . . . . . . 16,750 c . ...................................... iLj.6j"C. 49.5 . ')9. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . l 4 . W c. 97.09... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

11.95~ C.

. . .2. . .~. . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9,400c. In Thorpe's Dictionary of -1pplied Chemistry aiid also i n Rudorff's paper i n the Pharmaceutical Journal and Transactions the melting point given for acid of 99.5 per cent. strength is 16.6j0, but this appears to have been 3 misprint. A great deal of work had been done with this process before it was understood that t h e strength of the acid had any serious effect upon t h e results ; nunierous lots of acid had been obtained and used which as the event proved must all have been of approxiniately the same strength ; by ~

~

,

Xericlite, 3, 390, aiitl Phnrni-ccutical Journal aiitl Tr-nus~ictioiis.[j] 2 , 1 1 ~ Xeils!ein, 0 3'1 Ed., 1. 399-40s.

241.

S H EL.I A C A S A LYSIS

1325

comparing the results previously obtained upon standard sanip!es of shellac, using acid of untested strength, with those later obtained with acids of known melting point, it appears that the mass of valuable data which had been obtained, largely by Dr. Langmuir, concerning the iodine figures of pure shellac had been obtaiued with acid having a melting point of approxiinately 14.8'. We have, therefore, adopted as the staudard strength of the acid used in this process the strength which corresponds to a meltiug point of 14.7' to 15'. In order to ascertain quantitatively what are the effects of using acids of different strength, we have made several series of experiments in which the strength of the acid was the only variable. One such series was made upon a lot of shellac which had previously been tested a great many times and had always given a figure of approximately 1 7 . 7 ' . A quantity of Mrijs solution was prepared, using acid having a melting point of 14.95'~and with this Wijs solution three tests were made upon the same shellac, using in each test for dissolving the shellac acetic acid of different strength ; acid of I 1.80' M. P. gave an iodine figure of 14.67, acid of 14.95' &I. P. (the same a s that used for the preparation of the Wijs solution) gave an iodine figure of 17.59, while acid of 16.40' M. P. gave an iodine figure of 20.48. We have satisfied ourselves that equally correct determinations of the proportion of shellac and rosin in a mixture can be made by the Hubl, the Hanus, the Wijs and the bromine methods, providing that in each case pains are taken to examine a considerable number of pure shellacs and of rosins to form a satisfactory basis of calculation. Of these methods the bromine method is the simplest, as the standard solution is very simply made by adding a given volume of bromine to a given volume of glacial acetic acid T h e Hanus solution comes next in simplicity, but we are of the opinion that if sufficient care be taken in the preparation of the iodine broniide solution to insure the exact relation of iodine and bromine to each other that is necessary, there is little to choose in this respect between the Hanus and b'ijs methods. T h e reliability of any of these methods depends largely upon the number of determinations which have been made by it upon samples of shellac and of rosin which are known to be pure. In this respect we believe the Wijs method to be very decidedly preferable, on account of the number of data obtained by Dr. Langmuir by this method, and using the particular conditions advocated in his paper already referred to. W e have carefully considered the merits of each of these conditions, the weight, volume, time, and temperature, and are of the opinion that they can not for general m e be modified with advantage.

P A R K E R C. 3 I C I I . € I I X E T , E T . A I

1226

I n order to test practically the degree of accuracy and of regularit), to be expected from the method. we caused to be prepared by Dr. Laiigniuir three samples of artificial mixtures of shellac and rosin, designated recoritainiiig j per cent. of rosin : ' . B." containing g spectively as per cent., and C." coritai~iing13.j per cent. T h e w samples were distributed to the members of the conitnittee and aiialyzed, the true coinposition remaining unknown to those making tlie test uiitil the work had been completed. T w o members analyzed the samples by the method here recommended and two others by the Hub1 method. T h e results were as follows : Saiiii,le ".%"

.iiiai~st. I I

>retiion

Coilmittre HCbl

per c c n t . rosiii. + .s

1T '

:.I5

>ample "11 ' Der ceiit rosiii.

;.S 7.65

Sample " C " per cent. rosin. 12.2

13.27

ACTION O F BARIUM AND H Y D R O G E N PEROXIDES

I 227

Tlie use of the figure 2 0 0 instead of 2 2 8 as the iodine figure of rosin by his method will result in giving the true percentage of rosin with a little yeater accuracl- i i i most cases ; we believe, however. that in order to ;uard against the rather remote possibility that the use of the figure 2 0 0 night give n percentage of rosin slightly too high, it is better policy to .ise the figure 2 2 s . r H E ACTION OF BARIUM PEROXIDE AND HYDROGEN PEROXIDE UPON FORMALDEHYDE. By C. A L L A X LYFORD. Received J u l i e 15, 1907.

Tlie problem of the action of peroxides upon solutions of formaldehyde has been left in a rather unsettled condition. T h e present research h a s for its purpose the clearing up of this question. TT7hile experimenting tipon the oxidation of formaldehyde it was believed that by using bariuni peroxide there niiglit result mutual oxidation and reduction with the production of barium formate. Upon looking u p the literature of the subject t h e writer found that Geisow' believes to have obtained onlp barium carbonate and hydrogen from the interaction of these t v o substances. H e sal's he was unable to detect formic acid as an intermediate product either in neutral or acid solution. Geisow gives the following reaction : HCHO BaO, = BaCO,,-- H,. So far as the writer can ascertain, this work has neither been confirmed nor disproved by other writers. F o r a numher of reasons this reaction does not seem probable. I n the first place, acetaldehyde forms a hard resin with bari.um peroxide, no methane being evolved. If there were such an easy direct removal of the carbonyl group one n~ouldnot expect a resin to be formed. Geisow states that the dioxides of lead and niaugatiese acting upon f o r m a l d e h ~ d ein neutral solution produce formates and hydrogen. I t is known that uranium peroxide acting upon formaldehyde in neutral solution produces uranyl formate and hydrogen. PbO, z H C H O = (HCOO), Pb -- H,. LO, 2 H C H O = (HCOO), UO, H,. In alkalitle solutions it has been shown that hydrogen peroxide acts upon formaldehyde according to the following equation' : 2HCHO zSaOH H,O, = 2HCOONa H ? + 2H,O, (incold), from which it would follow that the reaction with sodium peroxide and sufficiently dilute formaldehyde would he : 2 H C H O 4- Na,O, = ZHCOONa H,. If 30 per cent. formalin be dropped upon sodium peroxide, there is,

-+

+ +

+

+

+

+

+

Geisow, Ber., 37, 515. Blank a n d Finkenbeiner, Ber., 31, 2979.