In his paper of 1887, the main outlines of which were given in
_Nature_ at the time (1887, vol. xxxvii., p. 179), M. Moissan showed
that pure hydrofluoric acid readily dissolves the double fluoride of
potassium and hydrogen, and that the liquid thus obtained is a good
conductor of electricity, rendering electrolysis possible. It will be
remembered that, by passing a strong current of electricity through
this liquid contained in a platinum apparatus, free gaseous fluorine
was obtained at the positive pole and hydrogen at the negative pole.
The amount of hydrofluoric acid employed in these earlier experiments
was about fifteen grms., about six grms. of hydrogen potassium
fluoride, HF.KF, being added in order to render it a conductor. Since
the publication of that memoir a much larger apparatus has been
constructed, in order to obtain the gas in greater quantity for the
study of its reactions, and important additions have been made, by
means of which the fluorine is delivered in a pure state, free from
admixed vapor of the very volatile hydrofluoric acid. As much as a
hundred cubic centimeters of hydrofluoric acid, together with twenty
grms. of the dissolved double fluoride, are submitted to electrolysis
in this new apparatus, and upward of four liters of pure fluorine is
delivered by it per hour.
This improved form of the apparatus is shown in the accompanying
figure (Fig. 1), which is reproduced from the memoir of M. Moissan. It
consists essentially of two parts--the electrolysis apparatus and the
purifying vessels. The electrolysis apparatus, a sectional view of
which is given in Fig. 2, is similar in form to that described in the
paper of 1887, but much larger.
The U-tube of platinum has a capacity of 160 c.c. It is fitted with
two lateral delivery tubes of platinum, as in the earlier form, and
with stoppers of fluorspar, F, inserted in cylinders of platinum, _p_,
carrying screw threads, which engage with similar threads upon the
interior surfaces of the limbs of the U-tube. A key of brass, E,
serves to screw or unscrew the stoppers, and between the flange of
each stopper and the top of each branch of the U-tube a ring of lead
is compressed, by which means hermetic closing is effected. These
fluorspar stoppers, which are covered with a coating of gum lac during
the electrolysis, carry the electrode rods, _t_, which are thus
perfectly insulated. M. Moissan now employs electrodes of pure
platinum instead of irido-platinum, and the interior end of each is
thickened into a club shape in order the longer to withstand
corrosion. The apparatus is immersed during the electrolysis in a bath
of liquid methyl chloride, maintained in tranquil ebullition at -23°.
In order to preserve the methyl chloride as long as possible, the
cylinder containing it is placed in an outer glass cylinder containing
fragments of calcium chloride; by this means it is surrounded with a
layer of dry air, a bad conductor of heat.
The purifying vessels are three in number. The first consists of a
platinum spiral worm-tube of about 40 c.c. capacity, immersed also in
a bath of liquid methyl chloride, maintained at as low a temperature
as possible, about -50°. As hydrofluoric acid boils at 19.5°
(Moissan), almost the whole of the vapor of this substance which is
carried away in the stream of issuing fluorine is condensed and
retained at the bottom of the worm. To remove the last traces of
hydrofluoric acid, advantage is taken of the fact that fused sodium
fluoride combines with the free acid with great energy to form the
double fluoride HF.NaF. Sodium fluoride also possesses the advantage
of not attracting moisture. After traversing the worm condenser,
therefore, the fluorine is caused to pass through two platinum tubes
filled with fragments of fused sodium fluoride, from which it issues
in an almost perfect state of purity. The junctions between the
various parts of the apparatus are effected by means of screw joints,
between the nuts and flanges of which collars of lead are compressed.
During the electrolysis these leaden collars become, where exposed to
the gaseous fluorine, rapidly converted into lead fluoride, which
being greater in bulk causes the joints to become hermetically sealed.
In order to effect the electrolysis, twenty-six to twenty-eight Bunsen
elements are employed, arranged in series. An ampere meter and a
commutator are introduced between the battery and the electrolysis
apparatus; the former affording an excellent indication of the
progress of the electrolysis.
[Illustration: FIG. 1.--FLUORINE APPARATUS.]
As the U-tube contains far more hydrofluoric acid than can be used in
one day, each lateral delivery tube is fitted with a metallic screw
stopper, so that the experiments may be discontinued at any time, and
the apparatus closed. The whole electrolysis vessel is then placed
under a glass bell jar containing dry air, and kept in a refrigerator
until again required for use. In this way it may be preserved full of
acid for several weeks, ready at any time for the preparation of the
gas. Considerable care requires to be exercised not to admit the vapor
of methyl chloride into the U-tube, as otherwise violent detonations
are liable to occur. When the liquid methyl chloride is being
introduced into the cylinder, the whole apparatus becomes surrounded
with an atmosphere of its vapor, and as the platinum U-tube is at the
same instant suddenly cooled the vapor is liable to enter by the
abducting tubes. Consequently, as soon as the current is allowed to
pass and fluorine is liberated within the U-tube, an explosion occurs.
Fluorine instantly decomposes methyl chloride, with production of
flame and formation of fluorides of hydrogen and carbon, liberation of
chlorine, and occasionally deposition of carbon. In order to avoid
this unpleasant occurrence, when the methyl chloride is being
introduced the ends of the lateral delivery tubes are attached to long
lengths of caoutchoue tubing, supplied at their ends with calcium
chloride drying tubes, so as to convey dry air from outside the
atmosphere of methyl chloride vapor. If great care is taken to obtain
the minimum temperature, this difficulty may be even more simply
overcome by employing a mixture of well pounded ice and salt instead
of methyl chloride; but there is the counterbalancing disadvantage to
be considered, that such a cooling bath requires much more frequent
renewal.
[Illustration: FIG. 2.]