36
FREILING, CROCKER, AND ADAMS
K,O, and H,O). Rather than treat various soil Samples chosen at ran-
dom, we have studied the effects of representative individual constitu
ents and pairs of constituents and their contribution to the overall
affinity for rubidium when present inasoilsample. This approach greatly
reduces the number of measurements to be made but has some fundamental drawbacks.
One kind of measurement consisted in equilibrating constituents
with vapor from a Rb,Q-TiO, source at 1272°C in a platinum isopies-
tic chamber. Under these conditions, cf was determined by transpiration measurements to be 1.1 x 107° mole/liter, assuming Rb,O to be
the vapor species. The actual species is unknown. Constituents studied
were SiO,, Al,O3, iron oxide,* and CaO. An attempt to include Na,O and
K,0 resulted in their complete replacement by Rb,O; therefore Rb,O
played a double role. It represented both the fission-product species of
interest and the displaced Na,O and K,O. The results are summarized
schematically in Fig. 15. This diagram shows the N,; value of each
constituent of a ternary equilibrium mixture. Neither MgO nor CaO
(the principal constituent of molten debris from a coral~-surface burst)
shows any affinity for rubidium. The highest affinity is shown by Al,O,
(major constituents of air-burst debris). The pair SiO,—MgO forms an
almost ideal solution; i.e., it takes up Rb,O in proportion to the SiO,
concentration, the MgO acting as an inert diluent. On the other hand,
CaO monopolizes SiO, and prevents it from taking un rubidium until its
concentration drops to the value represented by the compound 2Ca0:
SiO,. This compound behaves ideally with respect to SiO, at the remaining compositions. The system SiO,—Al,O, splits into two ideal
portions at the composition Rb,O° Al,O;°2Si0,. The systems Al,03;—
CaO and Fe,O,;,—CaO (important for tower bursts on coral surfaces)
show no ideality.
For application of these data to actual soil samples, a sample of
Silicate soil (Ambrose clay loam) of known compositon was melted in
the isopiestic apparatus and its uptake of Rb,O measured. By means of
the experimental data described above, it was estimated that the soil
would take up about 52% by weight of Rb.O. The measured uptake was
49%. A similar experiment was tried using asample of the rockforming
mineral orthoclase feldspar. The predicted uptake was 48%. The actual
uptake was only 32 to 43% with some uncertainty in the experimental
value, possibly because of failure to achieve equilibrium.
Results with MoO, vapor in the SiO0,-— CaO system show the opposite
behavior from Rb,O. Here SiO, takes up no MoO; whereas 1 mole of
*Regardless of whether the initial form of iron oxide is FeO, Fe,03, or
Fe,Q,, the species resulting from equilibration with air will be the same in each
case,