504 
PACIFIC SCIENCE, Vol. XX, October 1966 
dicular to the normal about which the waves are 
reflected. 
For crystals or particles below 20p in diame- 
ter, this requirement is no longer fulfilled and 
deviations from the Bragg equation become ap- 
parent. As the crystals or particles get smaller, 
the deviations become larger ; they are shown in 
X-ray diffraction patterns by an increasing 
broadening of the recorded reflections. Con- 
versely, the breadth of the reflections can be 
used as a measure of particle size. The methods 
by which this may be done are described in any 
standard text on X-ray diffraction methods 
(see Elenry, Lipson, and Wooster, 1951). The 
effect of time of crystallization on crystallite 
size may be studied by these methods. In this 
study 5-g portions of the wet Akaka soil were 
treated as follows: (a) With 0.4 molar potas- 
sium phosphate at pH 2. The reaction system 
was left to stand for 5, 54, and 200 days. The 
finer fraction of the solid phase was filtered, 
washed, and dried, as usual. The half-peak 
breadths of the 15.7 A peaks were measured 
and were plotted (log scale) against time, (b) 
With 0.5 molar potassium phosphate at pH 2 
for 6, 10, and 15 days. The preparations were 
washed and dried as before. The half-peak 
breadths of the 15.7 A peaks were plotted (log 
scale) against time of preparation. The results 
are shown in Figure 3. The crystal size of the 
taranakite increases with time, in accordance 
with expectation. 
PHOSPHO- REACTION PRODUCT 
PRODUCED FROM AKAKA SOIL 
TEMPERATURE «C 
Fig. 2. D.T.A. curves of soil-phosphate reaction 
product and pure synthesized taranakite. 
Fig. 3. Relationships between crystal size as mea- 
sured by half-peak breadth and time of preparation. 
DISCUSSION 
Clearly identifiable taranakite was produced 
under the conditions of the experiments; that 
is, with weights of soil varying from 5 to 15 
g, and with 0.2 to 0.6 molar solutions of 
potassium phosphate adjusted to pH 2. The 
pH values of the filtrate, and presumably of 
the reacting system, varied from pH 2.3 to 2.9. 
Many other concentrations of potassium phos- 
phate and other pH values were tried, but no- 
where outside and above that range was crystal- 
line taranakite detected. 
Chemical analysis gave K/P0 4 and A1/P0 4 
ratios close to those obtained with synthetic 
samples, although the results for A1/P0 4 did 
show that some contamination of the crystals 
with soil material had occurred. The contamina- 
tion did not cause significant differences to ap- 
pear between infrared patterns or differential 
thermograms of soil-derived and synthetic tar- 
anakites. 
The taranakite was probably formed by a 
series of reaction steps. The initial step would 
be the dissolution of Al 3 + ions from the 
amorphous, hydrous oxides, followed by the 
formation of a complex anion by reaction be- 
tween Al 3 + and H.,P0 4 ", and finally by precip- 
itation with K+ ions. Although the Al 3 + ions 
will be hydrated in solution, the infrared and 
differential thermal patterns indicate that water 
is not a structural constituent of the complex 
anion. 
No crystalline taranakite was detected at the 
pH values which are common to Akaka soils in 
the field. However, from the evidence obtained 
here and from a review of the literature, it is 
reasonable to suppose that a cryptocrystalline 
