Approximately 20 wt.% of the bauxite from Andoom in northern Queenslan
d, Australia is composed of material that cannot be accounted for by i
dentifiable well-crystallized phases. This poorly diffracting material
(PDM), found within the core of bauxitic pisoliths, has similar chara
cteristics to that of era-alumina (eta-Al2O3); a cubic form of alumina
. A differential XRD pattern of the PDM displayed a series of broad di
ffraction maxima attributed to eta-alumina with a mean crystal size of
9 nm. Unit cell refinement, on the basis of a cubic cell, gave a latt
ice parameter of a = 7.98 Angstrom for Andoom era-alumina. TEM and sel
ected-area electron diffraction revealed the PDM to be composed of min
ute (10 nm wide), randomly oriented crystals of era-alumina in close a
ssociation with Al-hematite. Chemical analysis using a nanoprobe showe
d Andoom era-alumina to be almost pure alumina with <2 M% Fe, <1 M% Si
and <1 M% Ti. The closely associated Al-hematite may contain as much
as 22 M% Al, however a value closer to the theoretical limit of 17 M%
is more likely. A broad absorption band at 3450 cm(-1) and 1630 cm(-1)
in the infra-red spectrum of the PDM indicates the presence of a subs
tantial quantity of H2O, strongly adsorbed onto the surface of the cry
stals. This is presumably due to eta-Al2O3's large surface area of app
roximately 2200 m(2)/g. The natural occurrence of eta-Al2O3, in bauxit
e may be the result of low H2O activities within the micro-environment
of pores at the time of crystallization. The epigenetic replacement o
f kaolinite with eta-Al2O3 and Al-hematite is put forward as an explan
ation for the formation of bauxitic pisoliths at Andoom.