SOME HOILISG-POIST CURT'ES' R Y J. K. 11.AYn-001>
Until recently, hiit little work has been done 011 tlie clistillation of perfectly miscible liquids. I n general, the probleni ha3 been attacked b!. measuring tlie vapor-pressiirt. with J.ai->.iiig coniposition, t h e teinperature heiiig kept con.itaiit. I n t l i i h iiianiier the characteristic plienoiiiena were ver!. coiiipletel!. drteriiiiiied by the work principally of E;oiio\valoff ,> the researclies of Roscoe' on solutions of halid acids in n-nter heiiig a1.w important in this connection. But vet.!. few pairs! lion.e\-er, have been stndied of €:yen the more coiiiiiioii liquids. T h e iiir-estigatioii here recorded \.\-as undertaken to acid to oiir knon-ledge of t h e properties of sucli licliiids tlieinsel\x,s anti in the hope that accumulated data will make possible a clexwinsight into tlie nature of the plieiioiiieiia preieiited. Iiistead iiieasariiig the x-apor pressures lvitli constniit teiiiperatul-e i t n.ould appear to be m3re profitable to adopt tlie pi-ocediire i i i ( ~ ~ iiearly like the coiiditioiis of actiial practice. ?'lierefore. a, I i x y receiitl?. lieen clone 11). Tlia!-er,' the variatioli of tlie lioiliiig-puiiit ivitli coiiipmitioii uiitlsr coilstant pressure ha:; liecii stiltlied. T l i t . apparatus nsed was e s w i t i d l y that of O~iiclcii-ff atid Caiiiero~i.~ Tlie tlieriiioiiieter n-as gi-atluated tc~) oiie-teiitli i 0.r ' and (mefiftieth could lie readil!. ehtiiiiatecl, biit i i i x-eiieral no attviiipt t(.) read closer tliaii one-tenth seemed to be \varraiitalilc, The tlieriiioiiieter had prm-iously been carefull!. ca1ihr:ated anti appreciable error could be occasioiied fro111 ail!. iiiacciirac!. of tlle instriiiiient. - i l l readings here given are iiiico T h e materials eiiiplo!-ed \.\-ere :ill carefiill!. piirifietl aii(1 i i i ~~
~~
i I
~-
Coiitrihutioii
froill
:lie Chr11iical Divibiuti, I-, S. I k l i i r t i i i v i i t of .i:ncui-
t Lll'(f .
\Vieti . \ i i i i . 1 4 3-1 ~ ISSI I , 1,iehig'h =\niialen. L I Z ) 7 1Sj9 1 : 116,x i 3 1860'. 1 Jour. Pliys. Clieni. 2 , : i S g S ; 3, j 6 rSgqi. ' :\iii. Clieiii. J n u r , 17, ,j; I 1S9j .
?'
.!
every case, before iisiiig, distilled through a high column still ; first and last portions being discarded. In no case did tlie variation in the boiling-point of the purified material amount to iiiore than 0.5' C and seldoni to more than 0.1"C. T h e alcoliol was distilled froiii lime iii the conventional way and afterm-arcls from aiihydrons copper sulfate. T h e iiiethj.1 alcohol after long staiicling over anhydrous copper sulfate \vas decanted and distilled. T h e acetone \vas long treated with calciuiii chlorid, filtered and distilled. T h e chloroforiii \vas n-ashed with water four times to get rid of traces of alcohol, carefully dried over calcium chlorid, filtered aiid distilled. IYitli tlie carbon tetrachlorid 110 treatinent \vas found iiecessarj- other than tlie careful distillation tlirougli a high coluiiin. ;111 the deteriiiiiiatioiis were iiiacle a t tlie prevailing atmospheric pi-essure, aiid the corresponding barometric readings are given. I n 110 case was the variation large enough to seriously affect tlie value of the experiments. I t is of course to be regretted, for piirposes of comparison. that these pressures could iiot he the sniiie in all cases, but iio great difficulty arises from this cause. T h e tabiilated results follow :
TABLEI .~lcolioland \rater ');,
.\lcohol
Teiiip
IIar
'j;,.~lcoliol
100.0
78.2>
-/ J- I . *
42.j
82.3j'
9j.9
78.4 79.2
751.4
37.4
83.0
jjI.7
80.35
7jI.S
30.9 24.4
-i J-. 1
62.6 54.7
j1.2 47.7
45.1
1
S0.9
I
S1.j 91.8
82.1
,
7j2.0 7j 2 . I
22.2
19.2
'I-S3- , . ?
j
00.0
j4.S
TABLE I1 Alcohol and ether I,, .\lcohol
Temp
Bar
',, ;\lcohol
'79.0"
7 7 2 ,I
49.2
56.4
b7.j
43.1
74.8 64.0
5S.j
7 72. I 7j 2 . 0
54.2
772.0
57.5 54.,3
49.5 47.3
7jl.5 771.4
100 0
36.4 25.I 11.8 00.0
Teinp Bar ~ _ _ _ jj+.,j ~
7j4.1
T.IBLE111
X 1coh 01 and carbo n tetra ch lo r i d '
Temp
.-llcoliol
Bar
78.97j.Sj 1
,
72.4 69.4
68.2 ,
6j.6 6 j .j 66.S
76S.S
768.2
;6S.S 769.0
'i6P.3 '7.3
i'68.g
11.4
768.9
h.9
1
i'6S.g
i68.3
i'6Y.4
66.22j
6j.j 65.6
tg5.4 $.O
5.3
65.h 66. j
2 .j
6s.5
768.4
0.0
77.2
$7.;
768.;
TABLE 11. C 11 lor of o r m a 11d carbo 11 tetra ch lo r i cl 1,
CHCI, 100. I
y2.2
79.9
[email protected] 62.3 j6.2 50.Y
20.j
'71.3
49.4
TABLE 1' Chloroform and methyl alcohol Bar - ~ _ _ _
'),
770.2 770.2 770.2 7 7 0 .2 7'70.2 7'70.2
CHCl,>
6S.4 62,S
j7.0 i I ~
j J . 1 j2.2
43.7
770.2
,751.;
770 4
25.0
770.4 7j0.4
12.8 0.0
765.4 767.9
TABLE1-1 Acetone aiid water 1,
-
.Icetone
Temp
100.0
j6.g3
Bar
~
772.5
59.1 60.7
772.6
6S.j 60 o
61.6
54.1
63.1
772.6 772.6 772.6
38.2
76.4
46.4
62.4 1
r r
64.2
ii 3 . 0
~
1 1
1
'j Acetone -
~
~
Temp _
45.2 40.5
64 5" 65.3
11.1
81.1
0.00
99 s
Bar _
~
I
'
_
_
-_ _
772.8 772 s 772.5 772.9 713.0 '37
1 i5.'
'772.8
TABLE1-11 Acetone and methyl alcohol
', .Icetone
Temp
~
'
Bar
(j .Icetone
Temp
Bar
765.2 765.2
j6.65" 55.85 55.9 j6.I 56.65 57.3 57.75
j6j.3
765.3 76j.3 76j.3 765.3
j8.0
111order to demonstrate more clearlj- the existence of a minimum point, the following determinatioiis were carefully made :
,:,Acetone
Temp
'$ .Icetotie
Bar
Temp
Bar
Temp
Bar
100.0
97.2
92.9
Table "111 Acetone and ether '1 .Icetone
~~
100.0
s4.4 71.7
6j.r 57.1 53 9 53,s
56.7j' 48.8
45.3 43.5
(;{, Acetone
Bar
Terrip ~
'
j6S.2 768.1 76S.1 768.0 76j.9
'~
1'
ll
~~
50.3 44.8
37.3 26.S 1;.6
40.3O 39.4
38.22 36.9
36.0 3 5.0
j69.2
-
.
Bar
772.5 772.j
71.7
-.5
--7
t
,
64.8
1
7j 2 . 6
772.6 772.7 772.7 7j2.8
Tlie existence of a iiiiiiiniuin point on this curve was more accuratelj- determined hy tlie following series, in n-hicli a Eeckiiiaiiii differential thermometer was used on account of the 1ar~:t. scale available in that instrument.
_;, .Icetone ~ ~ Teiiip _ _ 2.6.~'
100.0
91.3
'$ .Icetone
Ear ~
,
2.63
'
S I . ~I
2.62
,
772.6 7'72.7 -,12.j
i
'I'ciiip
13ar
71.1
- . 67"
62.3
--
2.S2
jX.1
2.93
.-
112.7
~
~
772.7 772.6
Tlie significance of these results is more rtdil!- seeii 111. reference to the accompanying charts. T h e temperatures nre plotted as ordinates, and the percentage coinpositions hj- TI-eiKlit are taken as abscissas. S o essential ad\-antage is to be obtained 1):. calculating tlie composition in relati\-e nuiiihcr of reacting weights, the cnrves being merely displaced to the side and the charts losing in c1earnes:j. For purposes of coiiipari.;on tlie resii;ts of Thaj-er are included. In Fig. I tlie results for inistiires of the \xrious liquids n-itli alcoliol are giveti. T h e curve for n-ater needs 110 comment, ai; its general properties were k n o \ m from tlie ~ v o r kof KcIno~vnloff. F o r the sake of completeness it \vas deeiiied wort11 wliilt. liox-e\-er to present these resalts. practical applicatioii, iii determining tlie composition of clilnte solutions of alcohol 111. their boiling-points, appears worth?. of consideratioil iii this C O I I iiection. Tlie cii-x-e n ~ l illustrates l the difficult!- i n f r a c t i o i i i i i ~ d c o h o l \vitli relativelj. w i a l l quantities of water. IYitli chloroforiii and carbon tetrachlorid respectiIdy \.t-r!- similar res111t,.;
Chlorofort~~
Fig.
I
\\-ere obtained. Both cur\ es show a minimum point. In the case of carbon tetraclilorid the mixture with constant boilingpoiiit contains about I .; j percent alcohol and boils a t about 65.j" C under a pressure of 768.4 111111. mercury. T h e slight waviness noticed by Thaj.er a t the alcohol end of the ci1rL.e for chloroform does not appear in the case of carbon tetraclilorid. 111 fact, if both curves had been determined at the
same pressure it seems eliclent that the cliloroforln one woiilcl rise slightl\- theti fall again until a t about a coniposition of 3 j percent cliloroforin i t wotiltl cut the cur1.e for carbon tetraclilorid aiid from this point 011 the two curves dil.erge fairl?. rapidlJ-. The curves for acetone and ether respecti\-ely present no 1111usual features aiid require no further coniment. -An experimental point of importance is to allow the contents of the apparatus to cool well before adding successive portions of ether. results here charted indicate the importance of this precaution, two of the boiling-points being oh\.ioiisl\. too high, probablj. on accoi.iiit of some of the material being lost b!, the violent elitillition resiilting from adding the ether to the hot alcoholic solution. I n Fig. 2 are plotted the ciirves for inistures in which
.-llcohol
Fig. z
chloroform is otie comtituent. IYitli carbon tetraclilorid the curl-e obtained has neither a iiiasiiiiiiiii nor iiiiniinutii point, and presents no remarkable features. The curve for iiietli\.l alcohol is i.ery similar in general features to the one for ethyl alcohol. I t slio\~--s a iiiiiiiiiiuni point corresponding to a mixture contaiiiing ahout 87.5 percent chloroforni and boiling a t 53" C under a pressure of j 70.2 miii of mercury. 'I'he waviness a t the inetli>.l alcohol end of the c i n e is quite as marked as in the case of
etliJ.1 alcohol. Indeed this waviness \vould appear to be a genera1 characteristic for curves corresponding to chloroforni solutions. Fig. 3 illustrates the results obtained for solutions in which
Sonit (2oiling--oiizf Ciii~i~cs
325
tlie alcohol and water ciir\.es respectii.ely woiild have intersected in t v o points. Up to mixtures containing about 30 percent of acetone the two cur\.es lie ver!. close together. KOiiiaxiniiiiii or niiniiiinni poiiit is s h o ~ v n . T h e ci1n.e for ether s h o w no special features. T h e ciin-e for n i e t h ~ - lalcohol, as i t appears on the chart, is verj- siiiiilar to that for ethyl alcohol. It, hmver-er, has a tnie iiiiniiiium point corresponding to 8;.or 88 percent x e tone, and a boiling-point of about 55.9" C a t a pressiire of 765 111111 of iiiercurj-. In other words, tlie boiling-point of the colistant boiling solution is only about 0.8" below that of the pure mlT-ent. I t is fortunatc:, in view of this fact, that we are not depetident upon fractional distillatioil for a separation of these siibstances. T h e c1in-e for carbon tetraclilorid is e i w i iiiore striking. It c1ot.s not in the least resemble that for chloroforiii. All niistiire: down to about 3 j percent of acetone have practicallJ- the saiiic boiling-point as the pure solvent, and fnrther the case is coniplicated by the existence of n true iiiinininni point, as was show1 b y careful esperiinents, correspoiidiiig to a concentration of 80 to 85 percent acetone and with a boiling-point otilj. about 0.0jc below that of the pure solvent. T h e curves for iiieth!-l alcohol and carbon tetrachlorid respectix-el!. lie strikingl!. near eacli other for solutions dowii to about 1 2 percent of acetone, and intersect each other twice. -1 curious coincidence is the intersection of the three curves for tneth!.l alcohol, carbon tetraclilorid, and cliloroforiii respectively a t practicall!. the same point . From the cases so far studied certain generalities are indicated, thongh the nuiii1)er of cases is hardly snfficient to justif!. iiiore than tentative statements. I t wonld seem that the existence of a maximum or a niiniiniiiii point is more probable nlien the boiling-points of the constituents themselves lie near one another. T h e case of carbon tetrachlorid and chloroform iminediatel5- suggests itself as an exception, but the fact of the c1o.e clieiiiical relationship of these two compounds ilia!. very readil!. be a strong inodifj.ing inflnence. From the results of Lehfeldt,' -
' Phil.
M a g . [j] 46, 42 (189s).
326
f.A: HayZiY7on'
it woiild appear that the curve for carbon tetrachlorid and benzene is also an exception, but i t cannot be admitted that the nature of this ciirve is established, on account of the verj- few observations upon which i t was plotted, and i t is not unlikely that further work will bring this case into line. TYitli the same solvent, substances of a similar chemical nature apparently j.ield similar curves. One well marked exceptioii appears : carbon tetraclilorid aiid chloroforni in acetone, for which no explanation is a t present obvioiis. For counsel and assistance in tlie inr-estigation here presented, I alii indebted to Dr. Fraiik I;.Cameron. A1tan early date I hope to present the results of observations upoii solutions in benzene, methyl alcohol and other solvents. From the facts liere presented the following conclusions may be clrawi : I . -111 mixtures of the following pairs of liquids boil at teniperatures betn-een the boiling-points of tlie constituents : alcoholit-ater, alcohol-ether, cliloroform-carbon tetraclilorid, acetoiiewater, and acetone-ether. 2. solution containing I 7.j percent alcohol in carbon tetraclilorid. distills without a change a t 6 j.j' C, approximately, under a presslire of 768.4 niin of niercury. 3. -1solution containing I 2 . 5 percent inethj-1 alcohol i n chloroforiii distills n-itliout cliange at 51" C, approximately, under a presslire of 770.2 111111 of inercurj-. solution containing I 2 to I 3 percent of niethj.1 alcohol 3. iii acetone distills n-ithout change a t 55.9' C, approximatel\-, under a pressure of 764.8 mi11 of iiiercur!-. j. -1 sohition containing 15 to 2 0 percent of carbon tetracliloricl in acetone distills without cliaiige a t a temperature but O . O j ' below that of the piire acetone, and all mixtures containing- more tliaii 40 percent of acetone boil within one degree of the boiling-point of the acetoiie itself. 6. T h e close proximity of the boiling-points of the con-
stitiieiits appears to he a favorable condition for tlie existence of a iiiasiiiiiiiii or a iiiiniiiiuiii point 1x1 the lloiliiig-piiit c1in-e. Similarity of constitution seeiiis to be a stroiigl). iiicxlif!.ing condition however. 7. Iii general, oiic constituent reiiiaiiiing tlie .same. iiiistnres with substances of similar clieiiiical coiistitution !-ieltl similnr boiling-point curves.