I.. ................ o.1464gram. 4 .................. 0.1464 3 ... - ACS Publications

The first paper that I know of which treats of the vapor-ten- sions of mixtures ...... Experience has taught me that about a liter an hour was about t...
0 downloads 0 Views 2MB Size
VAPOR-TENSIONS I..

OF MIXTURES OF VOLATILE LIQUIDS.

................ o.1464gram.

.................. 0.1464 3 ..................0.1460 2

4

615

.................. 0.1464

..

'( ' I

T h e separation of silver from copper and from cadmium is just as rapid and complete as these last separations. T h e deposits of mercury, gold, and silver, were carefully examined in the quantitative way for the various metals with which they had been associated; in every instance they showed themselves perfectly pure, so that these methods can be relied upon and trusted where accurate and rapid work is required. T h e metallic deposits were washed and dried in the manner described in previous articles. During the progress of the preceding experiments behaviors were observed pointing toward the separation of silver from gold, and mercury from gold and from silver in cyanide solution, but thus far expectations in these directions have not been realized. When conditions apparently favorable were obtained, traces of one or the other metal would be discovered in the metallic deposit, so that, a t this moment, trustworthy and definite data cannot be given. U N I V E R S I T S OF PENNSYLVANIA.

ON THE VAPOR-TENSIONS OF llIXTURES OF VOLATILE LIQUIDS. BY C. E. LINKBARGER. Received May 3. 1895.

CONTENTS.

Introductory. 2. Historical. 3. Description of apparatus. 4. Calculation of results. 5. Discussion of sources of error in apparatus. 6. Comparison of t h e vapor-tensions obtained by t h e method herein described and those obtained by o t h e r methods. 7. Choice and purification of liquids. 8. Preparation of t h e mixtures. 9. Experimental results with mixtures of normal liquids. IO. Influence of temperature on t h e vapor-tensious of mixtures of liquids. XI. Relations between t h e vapor-tensions, partial and total, and t h e concentrations of t h e liquid phases. I.

c.

616 12.

13.

14.

15. 16.

E . LINEBARGEK.

O N T H E VAPOR-TENSIOSS

Relations between t h e co!!centratioiis i n the qa.;cous aiid liqiiid phases. Heat of tlilutiori and solutioii; ideal coilcentrateti soiutiolls. T h e vapor-tensions of mixtures of acetic acid \vith I~euzeireand with toluene. The riiolecular mass of liquid acetic acid aut1 a general tnethoci of tleterriiining the molecular masses of liquids. ResurnC. I. I N T K O D U C T O K V .

T

H E investigation of the elastic forces or tensions of vapors

emitted by a solution of a fixed substance in a volatile liquid has received mucli attention, especially within recent years. T h e impetus for investigations of this kind is, in R great nieasure, due to the new riotioris that have been introduced into science in regard to the nature of solutions. T h e possibility of ascertaining the molecular mass of ;I substance from :i deterniination of the amount of the depression of the vapor-tension of a liquid, occasioned b y its being dissolved therein in know1 proportions, has induced chemists to study carefully this field of scientific inquiry, which it maj’ truly be said, lias been gone over very elaborately. I n the greater part of the work that has beer) done. both theoretical aiitl experimental, it lias 1,eeii assullied that tlie dissolved substance is not appreciably present in the gaseous state, and but sparingly present in the liquid state ; in other words, the dissolved substance is supposed to be involatile, aiid tlie solutions are made dilute. Now, absolute involatility in any body whatsoever cannot be affiriiied ; there niust always be, at every temperature, some degree of power of assuming the gaseous state, althougli it m a y be so slight as to be imperceptible to our senses. Still, for all practical purposes, the assumption of noli-volatility in many s u b stances can be admitted, as our means of experiiiietitation are not sufficiently delicate to detect any small amount of volatility. Although so much has beell done 011 the vapor-tensions of solutions of fixed substances in volatile liquids, coinparatively little attention has been paid to the study of the vapor-tensions of mixtures of the LroIatile liquids; yet this is the general case, of which the restriction that the dissolved substance be fixed

OF .MIXTURES OF VOLATILE LIQUIDS.

617

makes only a special application. I t must, indeed, be allowed that the consideration of a mixture of vapors, instead of a single one, introduces certain complications into the problem ; arid this is, perhaps, just the reason so little work has been done on this part of the subject ; still difficulties of this sort are probably not unsurmountable. T h e liniitations of work on vapor-tensions to dilute or, at most, moderately concentrated solutions cannot be said to be satisfactory. True, the theory of solutions has been developed on the hypothesis that dissolved matter, in analogy with gaseous matter, is in a state of considerable dilution ; and experimental confirmations of theoretical predictions can be expected only when such a state of affairs is realized. Notwithstanding that circumstance, it seems of importance to extend our line of operations and attack the problenis presented by concentrated solutions ; perhaps they will be found to exhibit fewer anomalies than has been supposed. There are two circumstances which render work that has hitherto been done on the vapor-tensions of mixtures of volatile liquids of all concentrations unsatisfactory ; they are to be found in the choice of the liquids investigated, and the kinds of vaportension measured. T h e liquids chosen were almost invariably those which ’are now recognized to be made up of associated molecules; they are just those which exhibit the greatest abnormalities in respect to niost of their properties, and it cannot be expected that simple relations, if they exist at all, will be discovered when such liquids are used as material of investigation. All investigators also, almost without exception, have measured only the total pressure of the mixtures of liquids examined, which is the sum of the partial pressures, these, however, being entirely unknown. But more important is it to know the share which each vapor has in the exerting of the total pressure, and only when this is learned can our knowledge of the matter be said to be in any adequate iiieasure complete. This paper seeks to fill in some degree this gap in the subject of vapor-tensions. T h e method employed is such as to permit of the specification of the partial pressures of a mixture’s components, and also of their concentrations in the gaseous phase. T h e

618

C. E. LINEBARGER.

ON THE VAPOR-TENSIOSS

choice of the liquids has been made with an eye towards employing those which have been found to he most " normal," so that in the examination of more complex liquids, that is, those consisting of associated molecules, the simplicity to he expected in the phenomena of the former may aid u s in getting some light on the possible intricacies of the latter. All the mixtures examined are freely soluble in one another so that no disturbing influence from layer-formation can take place. I11 reality, we have before u s R case of equilibrium ; the equilibrating systeiii consists of two substances, each present in two phases, the liquid and gaseous. We have to ascertain at the points of equilibrium the temperature, the partial pressures of the two substances i n gaseous phase, and their concentration in both liquid and gaseous phase. 2 . HISTORICAL.

T h e first paper that I know of which treats of the vapor-tensions of mixtures of liquids soluble in every proportion in one another is by Gustav Magnus,' who states that, when to a volatile liquid, such as ether, contained in a barometric vacuum, another less volatile liquid, such as alcohol, be added, the tension of the vapors of both liquids is less than that of the ether alone ; the cause of this behavior Magnus seeks in a certain reciprocal attraction on the part of the two liquids. Magiius' paper, being almost the first on the subject, contains, as, indeed, is generally the case with pioneer papers. some important statements, which, through the labors of later investigators, have become generalized into wide-reaching la thing in it is purely of a qualitative nature, no reliable quantitative data being given. Regnault' in the course of his extended investigations on the elastic forces of the vapors given off by liquids, determined a t 1 Ueber das Siedeu vou Gemeiigeu zweier Flussigkeiten uud uebrr das Stosseu solcher Gemenge. A n n . dev Phys. u. Chem. Pogf.. 38,481-492.2836. 2 L'eber die Elasticitatskrafte der Dampfe bei verschiedenen 'i'einperaturen itn Vacuo uiid in Gaseu ; uiid ueber die Spanuuiig der Dampfe aus gemengten oder geschichteten Phys. u. Chem. Pogg., 93, 537-579, zX54< and Memoires de 1' FliissiRkeiten ; A n n . Academie des Sciences, 16, 1862. Quatrieme Partie : Forces alastiques des Vapeurs qui sout ainises p a l les 1,iquids volatiles, Melangees par Dissolution reciproque on superposCs. Troisieine Classes : Melanges biiiairesdes Liquides qui s e dissolvent mutuellenieut en toutes Proportions, 724 aud 743.

OF MIXTURES OF VOLATILE LIQUIDS.

619

different temperatures by the static as well as dynamic method, the vapor-tensions of several mixtures of various liquids. Regnault did not pay much attention to the composition and analysis of the mixtures investigated, his object being merely to get a general idea of the relations of the vapor-tensions of the mixtures to those of the conipotent liquids. Still there is but a little doubt that the composition of the mixtures is specified with sufficient accuracy to admit of his data being regarded as reliable enough for theoretical considerations and confirmations. I n a later section are given his results so rearranged as to furnish some indications of the nature of the phenomenon which they represent. Regnault states in the papers printed in the MCmoires de 1’ Acadiniie that his experinients lead to the same conclusions as those of Magnus; but in the German translation (loc. cit.) from the Comptes rendus, he does not seem to have given Magnus this credit, which induced the latter in a paper’ immediately following that by Regnault, to call attention to his results published some eighteen years before-(loc. cit). I n this last paper by Magnus nothing new is communicated, a considerable part of it being occupied with quotations from his former pager. Plucker‘ determined by means of Geissler’s ‘ I Vaporinieter” the composition of the liquid and gaseous phases, as well as the total pressure of the latter, of a system consisting of a mixture of alcohol and water ; the work has the stamp of having been done with great care, and the accuracy of the results-rather meager, it niust be said-can probably be relied upon. Bussy and Buignet’ in the course of their researches on the physical properties of mixtures of hydrocyanic acid and water, made determinations according to the static method of the vaportensions of seven mixtures of the above two liquids (page 245 of their memoir). T h e work, which was carriad out at 13.25’ is fairly accurate, but the range of concentrations is not extellsire enough to permit of utilization of their results. 1 Ueber die Spannkraft der DBnipfe von Mischiingen zweier Fliissigkeiteri : .4 ) , ) i . dev Phys. 11. CJima. P o ~ g .93? . 579-582, 28.52. 2 Uiitersuchuugen uber DKinpfe uiid Diiiipfgeineiige : A n n . der P / i p , 21. Chon. P O X ,ga, ~ .193-220. ~ 2x54 : A continuntion of this article was promised by the aiithor, but I have been unable to find it aud regard its appearance as very doubtful. 8 Recherches sur lrtcide cyauhydrique : A n i i . CJiivi. P/iys., [4] 3, 231-263. 1664.

620

C . E . I,INBB,\RGEK.

O S THE Y . \ P O K - T I 2 S S I O X S

While the precediiig scientists for the most part endeavored to find relations ktn-eeii the tensions of vapors eniitted LJ- a iiiixture, and those gi\.eii off 1))- its coiiipoiients i n :I state of purity, I h c l a u x ' set himself about to get n knowledge of the relatioiis of tlie coiiipositioii of the liqnitl mixture to that oi the vapors emitted. His niethotl consisted in distilling :t large quaiitity ( I roo cc.) of a niixturc of krio\vii composition niid collecting several distillates, the coiiipositioii of which WIS tleteriiiiiietl through their surface teiisioiis b!. iiieaiis of tlie ' I3rop-iiiethod." Duclaux. believing, for rtmoiw which lie does iiot state, that simpler relatioiis are to be tliscovereti. i f the proportioiis of the liquids i i i a mixture be expressed i n \.olunies r:itlier thaii in weights, coiiimunicates results atitl tl:tt:t, which, : i b lie neglects to give any accurate indicatioiis as to tlic temperature at wliicli and tlie pressure uncler ivliicli tlie mixtures iii\wtigntetl b y lii til entered iiito ebullition, aiid as to the :Liiioiiiit ant1 tlirection of the chaiige of tenipcrature as the boiliiig proceeded, i: is iiiipossible to put into a shape periiiittiiig of comparisoii ivitli others ; the mixtures studied consisted of water nntl the series of alcohols up to caprylic nlcoliol, and of water ivitli foriiiic, acetic, : u i t l butyric acids. Wiilliier' deteriiiined according to the static iiiethod:' the traportension of five different mixtures of ethyl alcohol arid water at temperature intervals of about 10' from I I .8" to 84.6" ; also of two mixtures of sulpliuric ether arid alcohol at temperature intervals liad especially of about 3' froin 7.2' to a little over 30'. 117iill~ler i n iiiirid i n his work tlie determiiiatioii of tlie vai-iatioll of constancy of tlie ratio of tlie tension of the mixture of vapors to the suiiis of the tensions of each \.apor alone witli tlie teniperature : no inentioii is macle of experimental details and of the purity of tlie liquids experimented upon. It ma). be well fer the sake of completeness to iiiake niention 1 s u r 1t.s Forces 1:lastique k:rntiii~es p a r les M&lariges de d e i ~ aLiqiiides c J ~ I v ~ ,I%.i.r [ 5 ] 1 4 . .iO,i-.iJS. 1 ~ y i 8

.

11.

'

.-Inn

2"t7t.l,er die S p a i i l i l r n f t der I>iinipfc voii Fliissigl;eitsgetnischei~'': .4n71. d o . f'hys. (-hhrnr. f>o.q:r.) l a g , 3531.166. 1x66. 8Ann.

del. I ' / l v s .

16.

CJievii. F'(?,e.q , 103,5.i4-542

O F MIXTURES OF VOLATILE LIC,UIDS.

62 I

here of Alluard’s,’ Berthelot’s,* and Brown’s3work on the boiling points of mixtures of liquids, although little is to be found therein which bears directly upon our subject. An important experiniental as well as theoretical investigation on the subject in question has been made by Konowalow.‘ T h e method employed was the static, so arranged, however, that the error arising from the change of conipositioii of a liquid mixture due to the evaporation of its coiiiponents was reduced to a niinimum. T h e determinations were made with mixtures of water with each of the first four members of the series of alcohols, C, H,,+, 0,and of the series of acids CnH,,,+,0, at several different temperatures ; the work of Konoiralow may be looked upon as very accurate. Among Raoult’s nunierous and important publications on the vapor-tensions of solutions, there is one which deserves mention in the history of the vapor-tensions of mixtures of volatile liquids, even if one coinponelit of the binary iiiixtures investigated by him has but a feeble tension of vapor. T h e paper referred to’ treats of the vapor-tensions of solutions of turpentine, nitrobenzene, aniline, methyl salicylate, and ethyl benzoate, all almost non-volatile liquids, in ether ; the deterniinations were niade at ordinary teniperatures by the static method, and are to be considered as remarkably exact. In a later section, the data will be given in a modified form. I n the last five or six years papers by Planck, by LeChatelier, and by Neriist, treating of the theoretical side of the question, have appeared ; from their iniportance as well as for convenience of reference, their co,.tents will be quite fully reproduced. Max Planck’ enunciated certain relations between the differ1 ” Experieiices sur la Temperature d‘ Ebullition de quelques 316laiiges binaires de , Phjs..[4] I , Liquides que se dissolvelit niutuelletneut eu tous Proportious.” : A ~ i r i Chitir, 3%-392. 1894. 2 “Sur la Distillatioiides Liquides Melauges.” Coirrpt. vend., 57. 430(18631; A n i r . CJiiirr. P~?s.. [4] 1 , 3s4-392.1864. 8 ” On the Distillatioil of Mixtures of Carbon Disulphide and Carbou Tetrachloride” : Transactions of the Chemical Society of London. 39, 304. 1881. 4 “Ceber die Dainpfspannuugeti der Fliissigkeilgeniischeu.” : A W I . drr P1i.i~. IVlPd.. 14. 219, &I. 6 “L‘eber die Datiipfdrucke Atherischer Losuugeu.” : Ztschi: p//-vs. Chrm.. 2,353-j;3,

IW.

e “Ueber die Dnmpfspanuung von verdiiuuteu Losuugen fliichtiger Stoffe.” : Zlscht-. phjs. Clunr.. z,405-414.IS@.

622

C . E. L I N E B A K G E K .

OX T H E V A P O R - T I S N S I O N S

ence of coiicentrations of mixtures of two volatile substances in equilibrating gaseous and liquid phases, and the depression of tlie !yapor-tension, which permit of experitiieiital verification. Planck assullies the applicability of the law of Raoult-van't Hoff and that of Henry to the case of the vapor-tensions of mixtures of volatile substances (liquids); that is, there must be direct proportionality between lonering of vapor-tension aiid molecular concentration, and also between conceiitratioii in liquid phase and partial pressure in gaseous phase : furtheriiiore lie states expressly that his deductions are made for the case of dilute solutions only, and that the substances ill all phases of a system consist of normal molecules. Such a system made u p of a liquid and gaseous mixture in contact inay be represented by the symbol : ?zm, n,?n> f Z ' V L ' , ?iZ'7?i,', where ?z and ? E , represent numl)ers of molecules, and n z and v c , niolecular masses ; tlie accented letters refer to the vapor, and the unaccenteci to the liqiiicl ; tliose written with the subscript have refererice to tlie dissolved substance, those without subscript to the solvent; ?L aiid I / ' are large in comparison with ?L, and z 1 ' . T h e numerical concentrations of tlie individual substances are : , I1 )I ?f ' (-1 c = . .L. '=.-' ?I "I-

?Z+?L,'

.?1+11,'

C

-F I+ ? ;

I

?Zy-?Z]'~

If a reaction supervenes occasioning the following changes in the numbers of the niolecules : 612 : 671, : & ? & I : A?l,l = y : y , : y ' : ) I , ' , equilibrium occurs, in case this condition, y log c+y,logr,+y'logr'+y,'logc,'=logh', is fulfilled' ; ' A is a function of pressure arid temperature. I n the case before us, we have two diffeient reactions to consider: the vaporization of the solvent aiid that of the dissolved substance. Accordingly : y,=o y'=1 I . y=-I y,L=o y,=1 2.y=o y'=o y,'=1 T h e necessary conditions for equilibrium are : 1 S e e &I. Plauck. 4,S9 2x87.)

Ueber die Veririehrurig der Eutropie. A I I ~ Id?r , Phys. Ivied., 3a.

OF M I X T U R E S OF VOLATILE L I Q U I D S .

623

-logt+ log c’= log K

- log c, + log t,’= log K ,, or, if it be taken into consideration that t and t’ differ but little from unity, and if members of higher degrees be neglected : C, -c,’= log K, and log log K,.

/ z t , t / ti z n d Ch/cJr-u/b?+ill of . ~ + . ~ q l i l / i o ~5P~'vsioti iy- C t . i c = r . r / c , of .?+..'' i.i l.ij../) / / ) I / . i ! f .l[ci.crit:l,. 1 izflor-- Tt'tisiotz (g.Chlo~.oifo?-irrr z f .~,I..T' i s 289.2 7 i / t u . of A7/ci.ciLq,. hlolecules CHCI, in 1 0 0 Illole-

3loleciilcs CEiCI, i l l 100

illole-

ciiles of

'r c i i ~

ciiles of liquid mixtiire.

yaseoiis mixture.

16.97 j0.53 59.47

24.30

0.32.13

63.74 73.2j

1.1j1j

Grams CHCI:,

iii

(>r:ittis

w m of

C,ti, in

CHCl,

vapor.

vapor.

1.4770

iu

iniii.

0.660; 30.6 0.4187 13i".7 o.\ij31 162.2

'i't.11 sioii of

C..li,.

iii

iiiiii.

123.j

\'ulof

iiiii?

mr iii c c .

1032

74.3

IO~CI

50.2

ro,;o

1i:irviii-

1iltt.r-

etcr iii iiiiii

iinl

pressure 111 i i i i i i .

;jg 756 7.49

25

25

[TO 131; C O S T l S ~ ~ E l > . ] .- . - .-.

.

.

. . ....

[ C O S T K I D U T I O X IZKOlI T l I h Jotix I 1 1 I ( K I S O S 1,.1iiOK.\TOKJ' c l f l ~ : \ I I S T K Y , YO. 2.1

OF

T H E ELECTROLYTIC DETESiqINATION OF RUTHENIUil.

F

ROM tiiiie to time efforts liar-e lieeii iiixle i n tliis laboratory to gather iiiforiiiatioii upon tlitt tieportiiietit of the inctals of the platiiiuiii group torvartl tlie electric currelit. Pallncliuiii, platiiiuiii, aiid rhodiuiri Iia\.e I)ecii ~lt.teriniiietlqiiai;titativt.l!., and also separated electrolytically troiii other iiietals of the groul)? g.,plladiuiii from iridiuiii. 'I'lic 1iirrpose of this coiiiiiiiiiiication is to prcseiit data relntiii:,. t o tlic elcctrol!.sis of riitlieiiiiiiii salt solutions. Tlie literature of electrol!-sis cloes iiot coiitaiii an!iiiforiiiatioii upoii this point. Tlie salt upoii which the esperiiiients \\-ereinatlc rvas the double chloride of potassiuiii aiid rutlieiiiriiii. It \ v : i ~prepared b y fusing the finely divided nietnl with pntassiuiii iiitrate n i i t l li!.drositle. This fusioii was iiiatie i i i a silver crucihle. Tlie aqueoiis extract w a s acidified witli liytlrocliloric acid, anti the solution rvas tlieii evaporated to crystallization. Much potnssiuiii chloritlc s e p rated at first, but firially the double salt nppeareti i i i iiiiiiute redcolored iieedles. T h e platiiiuiii dish iii which the electrolytic tlecoiiipositioii \\.as carried out \vas coated upon its iiiiier surface with a la!.er of copper. Iii the first trials the solution of the double salt \vas mixed with three graiiis of sodiuiii acetate, a n d acted upon by a current of X . I).,&= o.oI-o.oj aiiipere. T h e quaiitit!. of tlie ruthenium salt iiot beiiig very abuiidaiit it \vas iiecessar). to coilduct the deteriiiitiatioiis with rather sinal1 aniouiits of material. (2,