Phase Diagrams for Ternary Liquid Systems

Rebecca J. Stead and Keith Stead

University of Exeter, Exeter, EX4 4QD. U.K.

Needing an undergraduate experiment on three-compo-

nent systems for students of engineering, we turned to the
.~
~~.~ ~" J. R. Clarke.' He advocated the use of the ternarv
naner hv
~

+
liquid system 1,2-dichloroethane (DCE) ethanoic acid
water and showed how the hinodal curve and tie lines could
he obtained by students in a normal three-hour practical
session. However. his method reauired the ~rovisionof cali-
bration curves of refractive indices for the analysis of the
mixtures a t the ends of the tie lines.
When planning to use this experiment i t became obvious
that a very simple adaptation of the original Clarke proce-
dure would allow the experiment t o stand alone without the
calibration curves.
In the first part of the experiment the position of the
binodal curve is determined. Starting with a known initial
mixture of the immiscible air of liauids water + DCE.
ethanoic acid is added from a buret &ti1 a homogeneous
solution is obtained. The innovation is simvlv to s a m d e this
homogeneous liquid and measure its refractive index. Not
onlv has a ~ o i n on
t the hinodal curve been fixed, hut its
refractive index is also known. From a series of such mea- +
Phase diagram tor DCE water 7 ethanoic acd st 23 "C and 756 mmhg. in
surements the hinodal curve can he drawn, and refractive part one, lhe nital oinary mixhlrss are r b w n ( X I with the tnai satdaled
indices can he added to the curve. Actual values vary ternary mixtures (A). In part two, lhe overall ternary mixtures are shown (t).
smoothly from 1.3330 for water to 1.4443 for DCE. wilh corresponding tie-line endpoints (a).
In the-second part of the experiment, ternary mixtures of
the liquids are made up and shaken in separating funnels.
The upper and lower layers are sampled, and refractive
indices determined. The overall composition can he calcu-
lated from the volumes used, and the compositions of the
two phases can he determined by linear interpolation be-
tween known points on the hinodal curve. A straight line Table 1. Blnodal Curve ~etermlnailon~
should join the two endpoints on the hinodal curve and the
v(H,0)/cm3 v(L?CE)/cm3 v(acld)/cm3 Macid) n
overall composition.
A mixture of DCE +water was shaken and then sampled; 10.0 8.0 22.6 54.3 1.3838
the two refractive indices showed the very limited solubility 6.0 11.2 18.2 49.0 1.3910
mentioned by Clarke.' I t would seem that no real inaccuracy 14.0 4.8 25.5 57.3 1.3754
is introduced if the mutual solubility is ignored and refrac- 3.0 13.6 12.5 39.6 1.4002
tive indices of pure components are used for the endpoints of 17.0 2.4 19.6 51.0 1.3684
the hinodal curve (also assumed to end a t the base apices of 1.0 15.2 7.8 29.2 1.4109
the trianele). 19.0 0.8 12.8 40.2 1.3594
The in% mixtures for the hinodal curve determination -1nitla1 ~ l u m s of
a water (HZO,denslty 1.00 g ~ r n - and
~ ) 1.2dichiommhane (DCE.
and the tie-line determinations together with typical experi- densny 1.24g~m-~)together wlmmevolumeof ethanoic acid(ac1d.dsnsihl 1.05 gem-')
mental results are given in Tables 1and 2 and in the figure. added to glve a homogeneous soi~lon, me mass peroem Weld) ot acid in the solution
and tha refractive index n of the ~ l u t i o nare given.
Solonn as the Abbe refradometer is calibrated with either
water ~ ~ D and C Ethe student keeps to one instrument, the
actual values obtained are self-consistent but may vary
slightly, with no detrimental effect, from the literature val-
ues.
The experiment has become free-standing and muchmore
satisfactory for the student; there is a real sense of achieve- Table 2. lnltlal Volumes for Tle-Llne Determination and
ment when the three points fit a straight line. Relracilve lndlces n 01 the UoDer and Lower Lavers

'Clarke, J. R. J. Chem. Educ. 1974, 51. 255-256

Volume 67 Number 5 May 1990 385