drawn; i.e., the excess oxygen atoms 
must occupy interstitial positions. 
This is confirmed by magnetic suscepti- 
bility and electrical conductivity meas- 
urements, and by a comparison of 
chemical, X-ray and densimetric 
analyses (8). 
Chemical stability. UO. powder is 
readily oxidized by air, even at room 
temperature. The extent of oxidation 
depends on the particle size and on the 
surface area exposed. 
According to Anderson et al. (9), 
room-temperature oxidation proceeds 
until the outer 50 A is oxidized, the 
oxygen absorbed being 0.8 cm* at NTP 
per square meter of surface exposed. 
Finely divided powders with particle 
sizes of 0.1 u, such as are obtained by 
the hydrogen reduction of ammonium 
diuranate at 900° C, oxidize to a 
composition approaching UOz..; after 
standing for one month in air at room 
temperature (1/0). Coarser powders 
(1 uw) are much more stable, oxidizing to 
only UOe.o2 under similar conditions 
(11). Once UO: powder has been sin- 
tered, at say 1,650° C in hydrogen, its 
oxidation rate becomes immeasurably 
slow. Thus, sintered UO, pellets may 
be stored for long periods in air at room 
temperature with no fear of oxidation. 
Sintered UOz is stable in deoxygen- 
ated, high-temperature water. Pellets 
showed only a slight dulling of their 
surfaces after more than 300 days of 
exposure in degassed water at 343° C 
or steam at 400° C at neutral or high 
pH (2). When 1-3 cm? of O2/kg was 
added to the water, however, a loose 
scale of hydrated oxide (UO;:0.8H20) 
formed, and the pellets lost ~15% of 
their weight after an exposure of 8 days 
at 343° C. 
In_ water-cooled power reactors, 
traces of hydrogen are likely to be pres- 
ent in the coolant stream due to the re- 
action of the water with structural 
metals. Thus, UOs should be stable 
in such systems. However, a steep 
temperature gradient will exist across 
the oxide radius, and if a hole or defect 
should develop in the cladding of a fuel 
element, the hot oxide core could be 
exposed to steam. Aronson’s calcula- 
tions (12) indicate that the core of such 
an element, operating near the melting 
point, could oxidize to UO2.13 to UO2,25. 
Direct evidence (13) indicates that 
stoichiometric UOz irradiated in a de- 
fective sheath may, in fact, increase in 
O/U ratio. A sample of oxide from a 
defected irradiated rod, initially stoi- 
60 
chiometric, was found to have a post- 
irradiation average composition of 
UOe.05. Recent Chalk River tests con- 
firmed this phenomenon. For exam- 
ple, a sample of oxide from the outer 
cooler region of a swaged, purposely 
defected rod changed from UOs..00 to 
UOz.2; during irradiation in a pressur- 
ized-water loop. As will be indicated 
later, an increase in O/U ratio may 
have a deleterious effect on the irradi- 
ation behavior of UO». 
The chemical stability of UO2 in 
other coolants has been studied. The 
oxidation rate of UOz in CO: has been 
measured at 500°-900° C (14). The 
UO. was much less reactive than U 
metal. At 700° C, the weight gain of 
sintered UO, pellets of density 9.6 
gm/cm* was 0.008 mg/cm?/hr com- 
pared with 400-560 mg/cm?/hr for U 
metal. Thus, there should be little 
Thermal Conductivity (w/cem/°C) 
201 209 217 
0/U Atom Retio 
FIG. 3. Effect of O/U ratio on thermal 
conductivity of UO». at 60° C (27). 
O—UOs, initial density 10.3 gm/cm%, 
sintered in Hy at 1,650° C, oxidized at 
900° C in 2-cm-Hg Os; A—UO, + 
0.1% TiOz, initial density 10.2 gm/cm’, 
sintered in Hy at 1,650° C, oxidized 
at 900° C in 2-cm-Hg Oo; O—UO,, 
density 10.6 gm/cm‘, sintered in steam 
at 1,400° C, cooled in steam 
worry about the compatibility of UO» 
with COs, unless there is an appreciable 
increase in the reaction rate in an irradi- 
ation field. 
Judging from thermodynamic data, 
UO: should also be stable in liquid 
sodium. There is some evidence (16) 
that high-density stoichiometric mate- 
rial is compatible with Na or NaK at 
600° C. 
Sintered UO, pellets are compatible 
with many sheathing materials. No 
appreciable reaction occurs between 
zirconium and UO, below 700° C (16). 
Little reaction occurs between UO, and 
TABLE Bernal Expansion of UO. 
Mean coeff. of 
Temp. range linear expansion 
(° C) (1058/° C) Ref. 
20— 720 11.5 34 
20— 946 10.8 5 
27— 400 9.0 
400— 800 11.0 35 
800-1,260 13.0 
400— 900 10 33 
aluminum below 500° C (17). No 
solid-state reaction has been found be- 
tween solid UO: specimens and beryl- 
lium or stainless steel at 600°-700° C 
(11). Sintered UO: pellets in contact 
with graphite plates do not react ap- 
preciably after a 10-hr heating in argon 
up to 1,500° C (18). 
Change in volume on melting. The 
limitation on central temperature in a 
UO: fuel element has not yet been 
established. Many designers have set 
the melting-point temperature as a 
limit until irradiation experiments yield 
experience on the use of elements with 
molten cores. Several values for the 
melting point of UOz have been pub- 
lished: 2,176° C (19), 2,500°-2,600° C 
(20), 2,878 + 22° C (21), 2,405 + 22°C 
(22), 2,760 + 30° C (23) and 2,860 + 
45° C (24). It seems reasonable to 
favor a value near 2,800° C as the true 
melting point. 
It would be useful to know if UO? ex- 
pands appreciably on melting to deter- 
mine what void space should be pro- 
vided in case the melting point is 
exceeded. Two measurements are un- 
der way in which UO, will be melted in 
a solar furnace (25) and a carbon-elec- 
trode are furnace with a water-cooled 
copper hearth (26). In both experi- 
ments, high-speed motion-picture 
cameras will record the solidification of 
a molten drop of UOs. 
Thermal conductivity. Unfortu- 
nately, UO has such a low thermal con- 
ductivity that the advantage of its high 
melting point is largely offset. How- 
ever, much work remains to be done 
before the thermal conductivity of 
operating UO» fuel elements can be pre- 
dicted with certainty, as indicated by 
the conflicting data plotted in Fig. 2. 
Many thermal-conductivity meas- 
urements have been made on sintered 
UOz in recent years (Fig. 2). Note 
that values differing by a factor of 
more than two have been obtained on 
