Second Method

Determinations of the relative composition of liquid and vapour by passing a known volume of air through a mixture at constant temperature have been carried out by Winkelmann,1 Linebarger,2 Gahl,3 and others.

In Linebarger's experiments a known volume of air (from 1 to 4 litres) was passed, at the rate of about 1 litre per hour, through the liquid mixture (40-80 grams) contained in a Mohr's potash apparatus, consisting of five small and two large bulbs. This apparatus was completely immersed in a suitable water bath, the temperature of which was kept constant within 0.05°.

When one of the components of the mixture was an acid, the amount of it in the vapour was estimated by absorption in potash or barium hydrate, and when one component contained sulphur or a halogen, the process of analysis by means of soda-lime was adopted. The total quantity of liquid evaporated, rarely more than 2 grams, was ascertained by weighing the bulbs containing the mixture before and after the experiment.

Preliminary determinations of the vapour pressures of a number of pure liquids were made in order to test the accuracy of the method ; the results were more satisfactory with the less volatile liquids - such as chlorobenzene, bromoben-zene and acetic acid - than with others, and they appear to be more accurate at low temperatures than at higher in the two cases in which such a comparison was made.

One pair of liquids investigated by Linebarger was benzene and carbon tetrachloride ; the results obtained appear improbable and they differ widely from those of Lehfeldt,2 of Zawidski,2 and of Young and Fortey,4 which are themselves in good agreement. It is to be feared, therefore, that for mixtures of volatile liquids the method cannot be regarded as satisfactory.

The tendency - so far as pure liquids are concerned - is apparently for the vapour pressures to be too low, and it may be that the air was not completely saturated with vapour, or that partial condensation occurred before the mixed air and vapour was analysed.

It is possible that the apparatus employed by Gahl (Fig. 29) would give better results.

For liquids of low vapour pressure, on the other hand, the method appears capable of affording accurate results. See, for example, Washburn and Heuse.5

1 Winkelmann, "On the Composition of the Vapour from Mixtures of Liquids," Wied. Ann., 1890, 39, 1.

2 Loc. cit.

3 Gahl, "Studies on the Theory of Vapour Pressure," Zeitschr. physik. Chem., 1900, 33, 179.

4 Young and Fortey, " The Vapour Pressures and Boiling Points of Mixed Liquids, Part II.," Trans. Chem. Soc, 1903, 83, 45.

5 Measurement of Vapour Pressure Lowering by the Air Saturation Method, J. Amer. Chem. Soc, 1915, 37, 309.

Fig. 29. - Gahl's apparatus.

Third Method

This method was first devised and employed by Rosanoff, Lamb and Breithut; the apparatus was afterwards improved by Rosanoff and Easley.1 The method depends on the principle that if a binary liquid whose components are in the ratio x/(l - x) is in equilibrium with a vapour containing the same components in the ratio p1/p2, then a saturated vapour of this composition will bubble through the liquid without producing or itself undergoing any change.

If the saturated vapour has the composition P1'/P2, the composition of the liquid will change until the ratio of its components has become x'/(1 _ x') corresponding to the composition P1 /P2 of the vapour. If a saturated vapour of definite composition at a temperature T be passed through a liquid mixture of the same components, the temperature T' of the liquid will gradually approach and finally reach T as equilibrium is attained.

The vapour bubbling through the liquid after attainment of equilibrium may be condensed in any desired quantity and analysed. Two consecutive samples should have identical composition.

Assuming that the ratio of the partial pressures in the vapour is equal to the molar ratio of the components, which is usually the case, and knowing the total pressure from direct observation, p1 and p2 can easily be calculated. Where the assumption is not admissible the relation between the molar ratio and the partial pressures can be determined by separate experiments.

A saturated mixed vapour of constant composition is passed through a liquid mixture of the same substances in a vessel a, Fig. 30, called the equilibrium chamber. It is surrounded by a second vessel B, in which the required vapour of constant composition is produced by boiling a liquid whose composition is kept constant by the introduction of the more volatile component in such quantity as to keep the temperature constant, as indicated by the sensitive thermometer c, the bulb of which is immersed in the liquid. The vapour of constant composition produced in b enters the equilibrium chamber near its bottom, bubbles through the liquid, then passes through a condenser, and the condensed liquid is collected in a receiver, like that shown in Fig. 27, from which consecutive samples of distillate can be withdrawn without interrupting the distillation or disturbing the total pressure in the apparatus. The delicate thermometer D in the equilibrium chamber indicates when true equilibrium has been attained, and this is further demonstrated by the identity in composition of consecutive samples of distillate.

The whole distilling apparatus is placed in a thermostat kept at a temperature slightly higher than that of the boiling mixed liquid.

The additional quantities of the more volatile component are introduced into B by regulated distillation of the liquid by electric heating from a vessel e provided with a condenser. Reservoirs F and G were also provided, from which fresh quantities of the mixtures (or of either component) could be introduced into A or b. There was a device to guard against the danger of liquid in A being carried as spray into the condenser. The form of the electric heater in b was not as shown in Fig. 30 but was of a conical spiral form so as to reach almost to the bottom of the tube and to ensure thorough admixture by bubbling. The arrangements for collecting successive portions of distillate, for regulating and keeping the total pressure constant, and other devices for avoiding errors, as well as the mode of manipulation, are fully described in the papers.

1 Log. cit.

Experiments were carried out with six pairs of liquids, some of which formed mixtures of constant boiling point. There was good agreement with Zawidski's results, and the authors conclude that Brown's results are fair but not so accurate as Zawidski's.

Fig. 30.