secondary cells with lithium anodes and immobilized fused_salt

u
I
,
l
u
I
I
I
103.
One experiment was done with a large excess of oxygen (F2/02 = 1/61.
The meacted F2 and O2 (21.6 moles out of an initial 28) were removed and
the reaction tube evacuated to < 5 p. The color of the solid remaining in the
sapphire tube was very dark brown. 'hen the liquid nitrogen bath was momentarily
removed and the sample was illuminated briefly with a flashlight, the
product detonated violently and shattered the sapphire tube. Whether the
detonation bias due to the presence of a large amount of F204 or to F20e is not
known. It is unlikely that the detonation was due to O3 since its vapor pressure
is well above 5 CI at 77°K and should have been pumped off. The very
unstable F20, (dark brown in color) has been reported to explode on illumination
or sudden warming. It is claimed to decompose thermally about 90°K.
In the experiments listed in Table 1, some F20 was formed in addition
to F202 and F204j the amount was relatively small and was not studied
systematically. The presence of F20 was detected by mass spectrmetric analysis.
Experiments Vith Added FFrl
It has been sham by others15) that BF3 reacts with F202 at low temperatures
to form the ionic campound O2%F4-. '!e therefore added BF3 to
irradiated mixtures of O2 and F2 in order to explore the radiation route to
O2+BF4- and possibly to 04+RF4-.
Boron trifluoride (3.5 moles) was condensed in the top of the
irradiation tube containing the formed fluorine peroxides (ca 1.7-2.0 millimoles)
suspended in the excess fluorine and oxygen. The BFs was distilled to the
bottom of the tube and the contents mixed by alternately vaporizing and condensing
a portion of the excess fluorine and oxygen. In this manner the BF3 was
better able to contact the reddish-brown peroxide. The excess fluorine and
oxygen were then removed under vacuum at 771. llhen the contents were warmed
to 113"K, a rapid reaction mcurred, and the color changed to orange. This
suggests that much of the F2O4 decomposed without reacting with BF3. The orange-\
color changed to white when the tem erature was raised to 143°K; this corresponds
to the conversion of the F202 to 02gF4-s Further warming to 2401 led to slight
deccmposition; the remaining solid was relatively stable, decomposing only very
slowly at rwm tercperature and atmospheric pressure. The decomposition is much
more rapid under reduced pressure, indicating a reversible step with a gaseous
product in the deccmposition. The yield vas measured by recovering the solid
in a dry box and weighing it.
The gases evolved upon warming to 240°K consisted of 02, F2, and BF3.
Irradiation of various fluorine-oxygen mixtures followed by addition of BF3
gave gases for this low-temperature decomposition in the ratio of (02 + Fz:/BFs
of 1.9 -1.35. There is a systematic tendency for this ratio to be higher when
the total volume of gas liberated is low. These findings correspond to
decomposition of an oxygen-rich canpound, probably 04+BF4-, with simultaneous
induced decomposition of some of the 02%F4-. (Evidence for the formation of
O4+~F4- from F2O4 has also been obtained in an independent investigation by
Soloman and his colleagues.)ls)
The more stable product (presumed to be O2+BF4-) decanposed rapidly
above 300°K to give 02, FBI and BF3. The elemental analysis of the solid was
F, 62.6 . B, 8.6$; theory, F, 63.7s; B, 9.3:). The infrared spectrum of the
po1:der %hieen silver chloride plates exhibits the characteristic absorption
frequencies?') of the BF4- ion. No absorption band attributable to 02+ was
observed, but this ion should have no dipole moment.