!
&
7
*
.
..
.
I
1
.
f
I OF I
ORNL
P.2796
.
24
1.
1
i
*
.
$
.
.
*
-
EEEEEEEE
J
11:25 114 11.
MICROCOPY RESOLUTION TEST CHART
NATIONAL BUREAU OF STANDARDS - 1963
*
.
.
-
ORIN -P-2496
Conf.670506.01.
1967
MASTER
FEB 1
PREPARATION OF REACTOR FUELS BY SOL-GEL PROCESSES*
CFSD PRICES
P. . Haas, F. G. Kitts, and H. Beutler
Oak Ridge National Laboratory
Oak Ridge, Tennessee
40. 63.0; wx_65
H.C.
LEGAL NOTICE
This report was prepared as an account of Govorament sponsored work. Neither the United
States, nor the Commission, nor any person acting on behalf of the Commission:
A. Makes any varranty or representation, expressed or implied, with respect to the accu-
racy, completeno88, or usefulaers of the information ccatained in this report, or that the use
of any information, apparatua, method, or process disclosed in this roport may not infringe
privately cwned rights; or
B. Asmumos any llabilities with respect to the use of, or for damages resulting from the
180 of any inforgiation, apparatus, method, or process disclosed la this report.
As used in the above, "person acting on behalf of Ka Commission" includes any em-
ployee or contractor of the Commission, or employee of such contractor, to the extent that
such employee or contractor of the Commission, or empioyee of such contractor prepares,
disseminates, or provides access to, any information pursuant to his employment or contract
with the Commission, or his employment with our contractor.

RELEASED FOR ANNOUNCEMENT
II NUCLEAR SCIENCE ABSTRACTS
For presentation at the American Institute of Chemical Engineer's National
Meeting at Salt Lake City, Utah, May 21-24, 1967, and for publication in
"Nuclear Engineering Series" by the American Institute of Chemical Engineers.
*Research sponsored by the U. S., Atomic Energy Commission under contract
with the Union Carbide Corporation
OAK RIDGE NATIONAL LABORATORY
Oak Ridge, Tennessee
UNION CARBIDE CORPORATION
: . for the
U.S. ATOMIC ENERGY COMMISSION
· PREPARATION OF REACTOR FUELS BY SOL-GEL PROCESSES*
.
P. A. Haas, F. , Kitts, and H. Beutler
Oak Ridge National Laboratory
Oak Ridge, Tennessee
ABSTRACT
The sol-gel processes were developed and applied at the Oak Ridge
National Laboratory for preparation of high density thoria and/or urania
fragments or microspheres. Continued development of these processes for
other oxides and for large-scale applications is reported here. Relatively
high strength, spherical particles of Th0.2, UO2, Pu02, and their binary.
mixtures were prepared with diameters of 50 to 700 microns and densities
of 90 to 100 per cent of the theoretical densities. Sols, dispersed as
drops in 2-ethyl-1-hexanol, are converted to gel spheres by extra. tion :
of water. Sol preparation and microsphere forming equipment presently
in operation have capacities of over 10, 1, and 0.1 kg/day for Thoz, VO2,
2)
and Pu02, respectively. Procedures for remote or shielded operation of
such equipment and improved apparatus for dispersion of sols into uniform
drops were tested. The plutonia sols are prepared from Pu(NO3), solution
by a precipitation-peptization technique. Sol-gel microspheres show
unusual drying and sintering characteristics. They lose up to 15% o.f .....
their original weight and show up to 34% linear shrinkage; yet unusually
high densities are obtained at relatively low temperatures. This behavior.
is being studied to determine mechanisms and better process flowsheet
conditions.
INTRODUCTION
The first sol-gel process developed and applied at the Oak Ridge
National Laboratory was for preparation of theoretically dense fragments
of thoria-urania for vibratory compaction into tubes (1)(2). The adaption:
of this process to preparation of spherical particles (3) and the exte
to other compositions (+), (5),(6) have also been reported. All of these
have the important advantages of simplicity of production and relatively .
. low temperatures for the final firing of the ceramic products. A sol-gel .
process will be described in general terms and flowsheets will be reviewed
as an introduction to the remainder of the paper.
.
.
Recent progress reported here includes results from continued develop-
ment of the process for preparation of small, spherical particles and
from application of the sol-gel processes for plutonia and plutonia mix-
tures. The unusual drying and sintering characteristics of the gels are
discussed.
*Research sponsored by the v. S. Atomic Energy Commission
under contract with the Union Carbide Corporation.
Sol Gel Process Description
Basically, our soi-gel processes consist of three major operations:
(a) preparing an aqueous sol, (b) removing water to give gel particles,
and (c) firing at controlled conditions to remove volatiles, sinter to
a high density, and cause any necessary reductions or chemical conversions.
The original process was developed for a thoria sol (1)(2). According to
the original flowsheet, thorium nitrate is steam-denitrated to thorium
oxide that can be dispersed into a stable sol by adding very dilute nitric
acid or uranyl nitrate solution. The sol is evaporated to form a gel,
and then fired to densify. After sizing, the oxide particles are suitable
for vibratory packing into fuel tubes.
Sol Preparation. - Sol preparation procedures were developed for most
of the metal oxides that might be used in irradiation specimens or fuel
elements. One purpose was to obtain the same advantages for these metal
oxides as are obtained in the thoria sol-gel process. Another purpose
is to be able to prepare mixed-oxide products by preparing mixed oxide
sols or mixing pure sóls.
The feed materials for our sol preparation procedures have been nitrate
salus or solutions. The products of many of the solvent extractio.2 pro- .
cesses are nitrate solutions and any residual nitrate remaining in the
sol or gel can be volatilized during calcination. Conversion of the
nitrate into an oxide sol requires the following four steps, which can
be combined or accomplished in different orders:
Pu(IV) are preferred for urania or plutonia sols.
2. Converting the metal nitrate into a hydrated oxide,
..
3.
Removing excess nitrate and nonvolatile impurities.
For instance, NH4NO2 or NaNO must be removed if the
hydrous oxides are precipitated by NH4OH or NaOH. ..
.
4. Dispersing the oxide as a stable sol.
The preparation procedure for thoria sols
is unique in that
the colloidal particle is formed by a steam denitration at temperature
approaching 500°C. For sols other than thoria, the colloidal particle
is formed in solutions or vet precipitates. Thoria-urania sols with
u Th ratios of less than 0.1 may be prepared by adding Vo, or uranyl
nitrate to thoria sols. The u(IV) is adsorbed on the surface of the
thoria particles.
The flowsheet for engineering studies of urania sol preparation is
shown in Figure 1. First, a uranous nitrate solution is prepared by the
reduction of a uranyl nitrate solution with hydrogen and a Pt catalyst.
The hydrous uranous oxide is formed by addition of NH4OH to a pH of 9.
The precipitated hydrous oxide is washed and is dispersed by adding HNO3
..
........
..
.
..
..
...
.
.
.
.
..
.........
...
.
.....6
W
AR
3
and HCOOH and then heating. The presence of formic acid helps to main-
tain a high (U(IV)/totač v] ratio and simplifies the precipitation and
washing (4).
Preparation procedures have also been developed for sols of oxides
of Pu, rare earths, Am, Cm, Zr, and the mixed oxides, VO2-Thoz, or V02-Zr02.
The plutonia sol-preparation flowsheet will be included in the section on
application of plutonia sol-gel processes. The other sols listed above
have been prepared in laboratory apparatus only and will not be discussed
further in this paper. .
º a
R'
L
":
4H-IM' M'
Gelatioii. - Sols can be gelled by a number of mechanisms including
removal of water, change in electrolyte concentrations, chemical reactions,
and temperature changes. Gelation by the removal of water minimizes the
difficulties of subsequenly drying and calcining the gel. The renoval ..
of water involves mass transfer by diffusion and cannot be as rapid as
gelation by chemical reactions or heat transfer. When gelation is by
the removal of water, the gel contains less volatile constituents than
the sol. If the gelation is by other mechanisms, all the volatile con-
stituents in the sol remain in the gel. This can make the drying, cal-
cining, and sintering difficult and may make them impossible...
Small, spherical particles are the preferred fuel shape for niany
reactor designs. Small, spherical gel particles (microspheres) can be more
made by forming drops of sol and gelling them by ex"raction of water into
ar. organic liquid (3). For thoria or thoria -urania microspheres, this
operation simply replaces the gelation by evaporation of water in the
original flowsheet. The process for forming microspheres will be reviewed .."
and recent developments described in detail.
.: Drying and Firing Drying and firing are necessary to remove volatile
constituents, to cause chemical reactions, and to sinter the paticles to
a high density. The temperature and the atmosphere are controlled while
the gel is heated to the sintering temperature and then cooled to room
temperature. Particles containing Voz or carbides must be protected from
oxygen. The drying and firing conditions were initially determined
empirically with little theoretical understanding of the mechanisnis.
cracking of particles, the densification during sintering, and the amount
of carbon and gases in the calcined product can vary greatly depending
on the conditions used. Therefore we have investigated these effects
and our progress in understanding them is reported.
PREPARATION OF MICROSPHERES
A process was developed for conversion of sols into spherical gel
beads of 10 to 1000 u diameter. Small spherical particles are the .
preferred fuel shape for use with pyrolytic carbon coatings or in dis-
persion fuels. In the microsphere preparation process, droplets of sol
are gelled by extraction of water into an organic liquid such as 2-ethyl-
1 -hexanol (2EH).
The process for conversion of a sol into gel microspheres involves
the following five operations:
1. Disperse the sol into droplets.
2. Suspend in an imuniscible liquid that will extract water
to cause gelation.
3. Separate the gel microspherès.
4. Recover the immiscible liquid for reuse... -
5. Dry the gel microspheres.
. .
-
The size of the product microsphere is determined in the first step. In
the second step, the extraction of water causes gelation and thus converts
the droplet of sol into a solid sphere. This is the key process step.
The interfacial tension holds the drop in a spherical shape. The maximum
droplet size is limited since very large drops will distort. A sur fac-
tant must be dissolved in the immiscible liquid to prevent coalescence
of the sol drops with each other, coalescence of the sol drops on the
vessel walis, and/or clustering together of partially dried drops. The
remaining three operations are simple in principle.
ao
We will report our recenc progress on (a ) operation of a microsphere
pilot plant which includes procedures and equipment required for a remote
facility, (B) development of dispersers for soi, and (c) drying gėl spheres.
The inicial development of the process for preparation of microspheres
and its application to VO2 or VO2-ZrO2 sols have been previously reported
(3)(7).
Microsphere Preparation Pilot Plant
A complete system for preparation of calcined oxide microspheres
from sols has been installed and operated as part of a Coated particle
Development Facility (CPDF) at the Oak Ridge National Laboratory. This
system has usually been operated with a feed rate of 25 cc/min of a thoria
sol or 1 kg of Thoa microspheres per hour with 2-ethyl-1-hexanol (2EH) as
the immiscible liquid. The capacity has been limited by the steam supply
to the distillation system and by the sol dispersion apparatus. This
microsphere column system is operated to prepare microspheres for other
parts of the fuel cycle program and to develop equipment and procedures
for remote operations of such a system.
The first four process operations listed are done continuously in
a tapered glass column (Figure 2). The sol is dispersed into droplets.
which are released into the enlarged top of the tapered column. These
droplets are suspended or fluidized by a recirculated, upflowing stream
of the 2EH. As the water is extracted and the droplets gel into solid
micrcspheres, the settling velocity increases. The column configuration
7
.ITA
KW
A
.
.
. -
-
.
.
.SC.13. *
*
12
.
.
om
1.1.
..
1
.
.
.
..
...
..
and the fluidizing flow rates are selected to permit the gelled particles
to drop out continuously while sol dioplets are formed in the top of the
column. Then the separation of the gel spheres froni the 2EH is completed
by discharging the product collector into a drier and draining the liquid
off through a woven wire cloth. The gel spheres are dried and calcined
batchwise. Fresh or purified 2EH is continuously added to the column
and displaces a stream of wet 2EH to a recovery system. Water is
removed from the 2EH by distillation:
In the CPDF microsphere column we have demonstrated long-term, stable
uperation with respect to the 2EH and surfactant. A single charge of
2EH was used for over six months. The coalescence and clustering problems
for the thoria drops were effectively prevented by about 0.2 vol %
Ethomeen S/15 and 0.05 Vol % Span 80 in the initial charge plus 10 cc
Ethomeen s/15 and 2.5 cc of Span 80 added per liter of thoria soi fed.
When the system was drained in preparation for equipment changes, the
only noticeable deterioration in the surfactant-2EH solution was cioudiness
from accumulation of very stilall gel particles.
-
The problems of remote operation have been only partly solved. Gel
spheres or calcined oxide spheres can be transferred hydraulically without
difficulty using 2EH or gas. The inventory of' sol drops or gel spheres
in the column is adequately indicated by average bed-density measurem
We still need simple, remote techniques for determining the optimum
fluidizing flow and for inspecting the gel spheres for dryness and absence
of clustering or distortion.
Sol Disperser Development
Many column operating difficulties would be minimized by more uniform
sol drop sizes; therefore, a variety of sol dispersion devices were
tested. Sol drops can be formed from a larger mass of sol by applying
one or more forces, such as gravity, centrifugal field, shear, inertia, .
interfacial tension, and electrostatic repulsion. To obtain uniform
drops and controlled diameters, both the force and the configuration of
the sol where the force is applied must be uniform and one or both must
be controllable. For all dispersers tested, the uniform configuration
is obtained by feeding the sol through orifices or capillarie
to 0.030 in. diameter. Among devices tested and the forces of importance
were:
1.
Two-fluid nozzles (3):
gravity.
interfacial tension, shear, and
2. Rotary feeder (3): shear and centrifugal.
3. Shear nozzle (3): shear and inertia.
4. Electrostatic. nozzle (8): electrical potential.
5. Falling drop dispersers: gravity.
...............
..a
..
......
.
..
.....
...
......
....
..
..
.
.
.
.
.
.
.
6. Vibrating capillary: inter facial tension, inertia,
and shear.
The two-fluid nozzles (Figure 3) have been the most useful sol dis-
persion devices as they are reliable, give good uniformity, and are easily
controllable over the whole sol drop size range of interest (200 to 2000 u
diameter). Single two-fluid nozzles consistently give 90 wt per cent of
the product within + 15% of the mean diameter Table 1). Multiple nozzles
for extended periods of operation have given 70 per cent within £ 10% of
a mean diameter. The sol drop diameter for a varicose type of breakup.
from interfacial tension (as shown in Figure 4) can be predicted by the
47
K a TV w. X
where
D = sol drop diameter
F = sol feed rate
Vmax = drive fluid velocity
ik = dimensionless constant of 2.25
For conditions where this equation would predict large sol drops (larger
than 800 u) the weight of the drop and the force from the drive fluid
become important and the above equation is not valid. The drop size
will then be uniform and reasonably reproducible, but it is a complex
function of the two flow rates, the nozzle dimensions, the interfacial
tension, and drive fluid viscosity.
The uniformity of sol drops from capillaries which are mechanically
connected to and vibrated by a loudspeaker (Figure 3) is better at
optimum conditions than from any other disperser. A simple, clean
vibration appears to be best, as secondary vibrations cause non-uniform
drops. The best results were for a continuous, approximately sinusoidal
liquid stream which breaks at the midpoint position with respect to
amplitude (Figure 5). "The amplitude for this type of operation was from
1/4 in. at 20 cps to 1/32 in. at 200 cps and was obtained by 1.5 to 4.0
volc inputs to commercial loudspeakers. Results for single c
are somewhat better than for multiple capillaries (Table 1).
The free-fall drop mechanism and the relationship between drop size,
orifice size, and interfacial tension are well known. The use of plastic
buckets with a large number of holes (Figure 3) provides a practical
capacity and avoids drop size variations from variable wetting of the
orifice by the sol. This disperser is useful for large drops only; the
orifice sizes necessary to produce drops smaller than 1000 diameter
are too small to be practical. The interfacial tension between the sol
and the ZEH varies with variations in the amounts of surfaceactive agents
................
......
.
....
present and thus causes variations in drop size.
Thoria Microsphere Drying and Firing
Conditions to avoid cracking of thoria gel microspheres during firing
wure investigated empirically. In general, the factors that minimize :
cracking are those that minimize composition gradients within the gel...
microspheres. The drying conditions were the important variables and :
the best drying conditions were superheated steam to a final drying
temperature of 200°C. The principal requirements for gel spheres from
regular thoria sols were as follows:
1.
.
The diameter is a principal variable with increased cracking
as the diameter increases. If the calcined products are
larger than 500 u diameter, the best conditions of super-
heated steam drying atmosphere to a final drying temperature
of 200°C are necessary to minimize cracking. For calcined
product diameters of less than 250 u, the best drying con-
ditions are not necessary. The amount of cracking for diam-
eters of 250 to 500 u is variable; the best drying conditions
can usually be compromised without. excessive cracking.
.
The amount of cracking decreases as the maximum drying
temperature increases from 100°C to 200°C.
MIN
3.
.
.
."
"
Atty.
The presence of superheated steam in the drying atmosphere
promotes the removal of. 2EH from the gel and reduces the
amount of cracking during firing. An air atmosphere during
drying may permit exothermic reaction while relatively large
amounts of 2EH remain on the gel and thus give very rapid
temperature rises and excessive cracking. Inert atmospheres
(Ar or Na) without steam remove 2EH rapidly only at tenpera-
tures of over 180°C as compared to rapid stripping at 120-
140°C by steam. ..
"
..
."T
.
I
MU
.
..
For thoria microspheres, the presently preferred conditions:
Drying
Ar atmosphere
Ar and steam atmosphere
25 to 110°C in 1 hour
110 to 200°C in 6 hours
..:
,
.
Firing
Air atmosphere
Air atmosphere
Air atmosphere
100°C to 500°C at 100°C/hr
500°C to 1150°C at 300°C/hr.
At 1150°C for 4 hours
A detailed and more theoretical investigation of the drying and
sintering characteristics will be reported as the final section of this
paper.
9.
1
TL
-
-
-
..
.
-
-
--
SOL-GEL PROCESSES FOR Pu022 PuO2-U02 AND Puuz-Tho2
: The development of plutonia solege i processes was prompted by the
earlier successes with thoria. The plutonia sols were converted to dense
fragments or microspheres by the procedures already developed for other
sols.
Puoz Sol Preparation
t.
+
Plutonia sol is formed by a precipitation-peptization flowsheet
(Figure 6). The batch size shown (Table 2) is convenient for the 4-ini.
diam geometrically safe equipment. A preliminary step is a valence
adjustment, if necessary, to ensure that the nitrate feed is in the Pu4
state. Bubbling NO gas through the Pu(NO3 )4 solution converts both Pub
and Pu3.7 to the desired state. A minimum of 1 M HNOS is maintained in
the feed to prevent polymerization; acidities up to 2.3 M were practical.
Excesses of base as low as 48 o proved satisfactory as long as the final
NH_OH molarity was 1 M or greater. The Pu(NO3)4 feed solution is added
to the NH OH solution at 10-20 cc/min with moderate agitation to insure
rapid neutralization and precipitation of the Pu(OH)4. The NH4NO3 formed
and the excess NH4OH are drawn off through a porous stainless steel (grade
G filter. The precipitate is washed thoroughly (5 washes with resuspen-
sion of the filter cake in each wash, . A high-nitrate sol is formed by:
adding HNO3 at a nitrate to Pu molar ratio of 2.5. A minimum ratio of i
is necessary for dispersion; ratios as high as 4 were used, but 2.5 is
sufficient to bring about dispersion upon heating to ~80°C. At this
point a true sol (crystallite size ~20 A exists, but microspheres
formed from this material would be of low strength and density. To form
desirable material the NO2 Pu ratio must be reduced to 0.17-0.11; this
is accomplished by thermal denitration, or baking. The sol is first
evaporated to dryness and heated to a temperature of 150-190°C.
The denitration reaction has an inverse time-temperature relation-
ship; equivalent results are obtained from 15 hours at 200°C, 2 hr at
250°C, or 15 min at 300°C. Heating must be uniform, and over-baking....
must be avoided as material reaching a NO Pu ratio less than ~ 0.1
cannot be resuspended by simple addition of water. Material in the 0.1
to 0.2 range on NO2/ Pu can be resuspended at ~ 1 M PuOz by agitation with
water. The sol may be concentrated by evaporation.
Batches containing 50 to 100 g Pu were precipitated using feeds
containing 41 to 115 g Pu/liter and NH_OH solutions from 2.5 to 6.3 M
(Table 2). The NH OH excess was 100% or 1 M in the final solution,
whichever was less. Dispersion was carried out at NO. /Pu ratios from
the practical minimum of 1.3 (requiring several hours at 80°C) to 2.5
(instantaneous at ^ 80°C) at concentrations from 20 to 53 g Pu/liter.
Final sols containing 1 to 2.5 M Puoz at NO3 Pu ratios of 0.1 to 0.3
were prepared.

Microsphere Forming
Plutonia and PuOz-Th02 microspheres are formed in a manner yery
similar to that used for Thoz microspheres. For mixtures, the two sols.
are prepared separately and blended to the desired proportions; thus,
any Pu/th ratio can be easily achieved... The drying solvent is 2-ethyl- **
Ethomeen S/15; small amounts of Span 80
can be added to control sticking to the wall if necessary. Plutonia
microspheres are dried to 150-170°C in argon; PuOz-Thoz is heated to
.~ 120°c, steam stripped, and then taken to 150-170°C ir argon. Both are
fired to 1200°C in air.
Plutonia-urania mixtures must be treated like urania; atmosphere
control is required until after firing. Again pure sols are prepared
and blended to any desired composition. The drying solvent is 2EH with
a surfactant system of 0.5 v/. Ethomeen S/15 -0.5 y/o Span 80. Drying
is 150-170°C in argon followed by firing to 1100°C in argon -4% H2 with
a CO2 cycle at intermediate temperatures to remove residual carbon.
Products and Evaluation
The compositions, size ranges, and some physical properties of three
types of material produced in our demonstration facility (three 6-ft a
boxes) are shown in Table 3. About 2 kg of 20% Pu02-V02 were produced in
filling a commitnient to ANL for irradiation testing as a fast reactor fuel. ::
A 500 g batch of 5% PuO-U02 was produced for fluorination studies at ORNL
in support of the Fiuidized Bed Vol.atility Pilot Plant. About one kg of
coarse and half a kg of fine 5% Fu02-Ih02 were made for fabrication of
irradiation specimens at ORNL. The densities of all material (Table 3)
were greater than 95% theoretical with PuOz density approaching 100%.
Surface areas were low and showed the expected variation with size.
Carbon content was routinely below the limit of detection. Gas release
was low except in the case of UO2-PuO2, which had been fired only to
1100ºC during calcination. Crushing strengths for the coarse spheres
were about 1 kg/sphere.
In a photograph of two size ranges of PuO2 microspheres (Figure 7)
all appear round with glossy surfaces. A few surface cracks can be seen
in the larger spheres in the larger size range. This is also seen in
PuO2-U02 bit not with PuO2-Th02. Cross sections of both sizes show sharp
edges, consistent with low surface areas (Figure 8). There are no inte
voids although a small amount of microporosity can be detected.
DRYING AND SINTERING OF SOL-GEL THORIA AND URANIA MICROSPHERES
.
Sol-gel microspheres as extracted from the spheroidizing column show
some unusual characteristics. They are less than 40% of final density,
contain large quantities of volatile compounds such as water, nitrates,
organic solvents and surfactants and yet have remarkable strength and are
extrenie ly resistant to mechanical abrasion. In course of drying and
'
'
sintering these microspheres lose up to 15% of their original weight and
suffer large dimensional changes (up to 34% linear shrinkage); yet,
unusually high densities (98% of theoretical density) are obtained at
tenperatures much below levels ordinarily required to densify convention.
ally derived powder agglomerates.
Drying
Most of the volatile compounds in the gels can be removed by heat
treatment either in air or inert atmosphere. The complete removal of
organic coumpounds remaining from the gelation and sphere forming process
requires special treatments discussed later. Physically bound compounds
can be driven off rapidly in vacuum or by heat treatment at atinospheric
pressure and temperatures between 100° and 140°C. No dimensional changes
can be detected during this operation and particles reexposed to moisture
maintain their integrity. The remainder of water, nitrates, and organic
compounds appears to be strongly bound and can only be removed at higher
"emperatures (500-600°C). Removal rates increase rapidly with temperature.
wintering of sol-gel ceramics, urania in particular, starts at very low
temperature (400-500°c). It is doubtful if volatile materials can be
completely removed within reasonable times without concurrent sintering:
The removal of the chemisorbed species is associated with a dimensional
change. In case of thoria microspheres, a consistent shrinkage of 4%. .
was measured; whereas, in the case of urania microspheres, the initial .
shrinkage ranges between 8 and 10%. M. W. Wadsworth and R. K. Chang
investigated the initial densification of. thoria gel (9) and found a
linear relationship between shrinkage and weight loss up to 500°C. After
the chemisorbed species have been removed, microspheres show extreme
affinity for water vapor and exposure to moisture causes release of energy
and results in severe cracking.
We detected energy changes during the drying process by differential :
thermal analyses (DTA) in which specimens are heated at a uniform rate
(10°C/min) in various atmospheres. A typical DTA pattern obtained during
drying of sol gel thoria fragments is shown in Figure 9-A. The two
pronounced endothermic peaks starting at 70°C and 300°C are attributed
to the release of water and nitrates.
Particular problems are assoicated with the removal of residual.
organic compounds. If the drying of thoria microspheres is carried ouc, ..
in air, rapid oxidation of organic compounds starting at 180°C leads to
severe temperature excursions in bulk charges of microspheres and causes
severe cracking. A similar problem, however, occuring at a slightly
higher temperature (starting at 230°C) was also noticed for heat treatments
in inert atmosphere for both thoria and urania microspheres. DTA analyses
of sol-gel microspheres (Figure 9-B) in helium showed a strong energy
release starting at 220°C. This particular energy release has been
attributed to an organic-nitrate oxidation reaction since it was only
noticed for gels containing organic compounds. The reaction appears to
end at 350°C. Subsequent treatments at high temperature for thoria fuel
.
.
.
A
can proceed in air or oxygen without difficulty. Efficient removal of
organic compounds to low levels prior to sintering appears necessary,
particularly in the case of UO2, to ensure low residual-carbon levels
ir. the sintered product. Since VO2 is unstable in tlie presence of free...
oxygen at these low drying temperatures (400-500°c), it tends to compete
in oxygen pickup with the carbon -oxidation reaction. Mild oxidizers
such as CO2 or water appear to minimize oxidation of 102. Steam was
extremely effective at low temperatures (120-140°C) for removing the bulk
of organic compound without oxidation therefore diminishing heat and gas
evolution caused by the organic-nitrate interaction. A significant ::...
reduction in energy release was noticed (Figure 9-c) after steam treatment
at 152°C for 8 hr prior to the inal drying operation. At elevated
temperatures (400-500°c), supereated steam is an efficient means to
remove decomposed organic-compound residues, probably by a water -gas-
reaction.
We determined the removal of organic compounds by steam from thoria
and urania microspheres as a function of time at 200 and 400°C (Figure
In case of thoria microspheres the residual carbon content could be
reduced from 6% to a level below 40 ppm within 1 hr using superheated
stean at 400°C. Also an immediate, drastic reduction of the residual
carbon from the same batch to 500 ppm occurred at 200°C within 15 min., :
prolonged treatments at this temperature did not result in further removal
of organic compounds. In case of 102 sol-gel microspheres containing
~ 8% residual carbon, the removal of carbon by steam was less effective."
The residual carbon content after a 6-hr treatment in steam at 400°C was
150 ppm. The effectiveness of carbon renoval at one particular time and
Cemperature appears also to depend upon the size of the spheres.
Sintering
All the oxide fuels derived by the described sol-gel techwique
exhibit extraordinary sintering characteristics. Densification occurs
-
can be obtained readily.
.
We have studied the concurrenc processes of densification and crys-
tallite growth in thoria and urania gels. Both processes cause a decrease
in svecific surface of the gel with an associated decrease in surface
enthalphy attributable to solid-vapor interfaces. We have detected this
heat release by differential thermal analysis (DTA) at uniform heating .
rates on both thoria and urania gels. In the case of thoria gel heated
in air, the heat release begins at 600°C and ends at 1200-1250°C. For
urania gel, the start of the energy release and the release rate both
depend strongly on the atmosphere. Calibration of the DTA apparatus (10)
enabled quantitative measurement of the total heat released and this .
yields a value of 1060 + 220 erg/cm² for surface enthalpy of the thoria
gel in the range of 600-1200°C. Predicted values for a (111) plane (the
lowest energy surface in thoria) range from 810 to 1090 erg/cm2.
.
The isothermal shrinkage kinetics of sol-gel thoria and UOą micro-...
spheres have been studied by sequence photography in a hot-stage micro-
scope. Experimental shrinkage isotherms were obtained for both thoria
and urania (Figures 11 and 12). Initial shrinkage due to a loss of
volatiles as discussed earlier is deducted. The isotherms differ from
those expected from imple sintering theory in that the log-log plots of
frictional shrinkage versus time are not linear but decrease is slope with
increasing shrinkage. The change in slope with increasing densification.
is attributed to the combined operation of a number of factors, including
the very small crystallite size, concurrent crystallite growth, the .
faceted crystallite morphology, and the possible presence of more than
one mechanism for material transport. Model equations (11) have been
set up to aid in the interpretation of data. Although still highly
simplified, they do overcome some of the objections to the direct appli-
cation of the conventional sintering equations. On this basis it appears
possible to eliminate volume diffusion as the rate-controlling mechanism
at shrinkage below 10% in thoria. On the basis of present theories and
experimental results, the sintering process seems to involve the removal:
of material from the inter-crystallite boundaries, migration along the
boundaries to the crystallite surfaces and then migration around the
surfaces away from the grain boundaries. It is not clear whether grain-
boundary or surface diffusion is the slower, and thus rate-controlling,
process. The surface diffusion process gives slightly better agreement
with experimental isotherms, whereas the equations suggest that grain-....
boundary diffusion should be the slower process. Use of the equation
for either process yields an activation energy of around 70. kcal/mole for
the relevant transport mechanism over the temperature range 650-900°C
and up to 10% linear shrinkage. This value is in agreement with one of
63 kcal/mole as obtained by observing the effect of sudden temperature
change on the shrinkage rate of thoria microspheres.
Changes in crystallite size and BET surface area have been determined
for thoria and urania during isothermal heat treatments. It was found
that x-ray crystallite size increased (using the broadening of 111, 220,
and 311 üeflections respectively) and the BET surface area decreased with
both time and temperature (Figure 13 The relationship appears to be
independent of temperature and this suggests a similar activation energy
for densification and crystallite growth. The isothermal crystallite
growth curves can be interpreted in terms of a very simple model involving
surface diffusion, which yields an activation energy of 61 kcal/mole in
case of thoria gel.
As expected in the case of urania, the sintering atmosphere greatly
effects the sintering kinetics. We have studied the sintering behavior
urania microspheres in "dry" and "wet" (Ar + 4% Hp) as
as in CO2 atmosphere. Our experimental data suggest that water and CO2
both enhance the sintering and crystallite growth. It appears, however,
that the relationship of shrinkage to crystallite growth and change in
BET surface area is still maintained. The addition of wa
+ 4% He gives an effect similar to a 50 to 100°C increase in temperature
that a steam atmosphere is extremely
.
-
:
13
effective in increasing the rate of sintering of UO2. Carbon dioxide
atmosphere also caused a drastic acceleration of the sintering process.
The surface energy release during DTA started 200°C lower in coz and was
completed within a much narrower temperature range compared to treatments
nt 4% hydrogen (Figure 14). Isothermal heat treatments also
confirmed that both sintering and crystallite growth were greatly enhanced
by coa atmosphere. So far, we have not been able to demonstrate whether
these effects are associated with oxidation of the urania during the
sintering process.
The uranium oxide used for our investigation had a high initial
oxygen content (0/0 = 2.281). We have not determined changes in composi-
tion during sintering. After final densification in dry or wet argon
+ 4% hydrogen atmosphere and cooldown in helium,' the final 0/U ratio was
less than 2.004. We expect that the initial uranium oxide composition
has a bearing on the sintering behavior. Studies to determine this
effect are still in progress.
ACKNOWLEDGMENT
This paper includes, or is dependent on, work done by a number of
persons at the Oak Ridge National Laboratory. Of particular importance
were the efforts of S. D. Clinton and C. C. Haws in preparing microspheres,
M. H. Lloyd in developing the Pu sol-gel flowsheet, and M. Bannister in
the drying and sintering of thoria gels. .
REFERENCES
1.
.
D. E. Ferguson, 0. C. Dean, and P. A. Haag, Preparacion of Oxide Fuels
and Vibratory Compaction by the Sol Gel Process, CEND-153 (Vol I)
p 23 -38, also ORNL-TM-53 (November 20, .961).
2. d. L. Lotta, et al., "The Oak Ridge National Laboratory Kllorod
Facility," TID-7.550, pp 351-383 (1962).
Paul A. Haas and S. D. Clinton, "Preparation of Thoria and Mixed Oxide
Microspheres," I and EC Product Res. Dev. 3 (3) 236-244 (1966).
J. P. McBride, "Preparation of loa Microsplieres by a Sol-Gel Technique, "
ORNL-3874 (1966).
5. J. L. Kelly, A. T. Kleinsteuber, et al., "Sol-Gel Process for Preparing
Spheroidal Particies of the Dicarbides of Thorium and Thorium-Uranium
Mixtures," Ind. Engr. Chem., 4, 212-216 (April 1965).
6. P. A. Haas, et al., "Sol-Gei Process Development and Microsphere
Preparation," 2nd Internacional Thorium Fuel Cycle Symposium,
Gatiiņburg, Tennesse, May 3-6, 1966.
7. P. A. Haas, s. D. Clinton, and A. T. Kleinsteuber, "Preparation of
Urania and Urania-Zirconia Microspheres by a Sol-Gel Process,
Canadian J. Chem. Engr. December, 1966.
8. D. M. Helton, "Dispersion of a Liquid Stream by an Electrical
Potential: Applications to the Preparation of Thoria Microspheres,"
U. S. AEC ORNL-TM-1395 (January 17, 1966).
9. M. E. Wadsworth, A Fundamental study of Thoria and Urania Gels,
Progress Report, January 1966, Research and Development Subcontract
2176 (University of Utah) under W-7405-eng-26.
10. M. J. Bannister (to be published, Journal of Physical Chemistry).
11. M. Ė. Wadsworth and A. M. Daniels, A Documental Study of Thoria and
Urania Gels, Progress Rept., January 1966, Research and Development
Subcontract 2176 (University of Utah) under W-7405-eng-26.
12. W. E. Baily, et al., "Steam Sintering of Uranium Dioxide," Am. Ceram.
Soc. Bulletin, 47 (11), 168-172 (1962).
Table 1. Product Sizes of Calcined Thoria Microspheres
From Three Dispersers
Sol feed:
Thoria sols of 3.0M Th, sol drop diameters of 2.35 times
diameters of theoretically dense Thoz product
Two-fluid Nozzles
Vibrating Capillaries
Free-Fall
Drop
Single
250
1.2
Six
425
Single
425
1.2
Four
480
19.2
200
Four
480
9.6
19
400
9.6
24.7
40
50
530
3106
3906
2308
10,200
314
720
480
• Number of feed capillaries
Capillary diameter, u
Sol feed rate, cc/min
Vibration frequency, cps
. Predicted mean size, u
Amount of sample, g
Weight per cent in:C
30/35 or 500-590 u
35/40 or 420-500 u
40/45 or 350-420 i
45/50 or 297-350 u
50/60 or 250-297 H
60/70 or 210-250 j
• 70 or < 210 H
97.9
0.8
1.1
0.2
99.3
1.5
و
0.1
61.9
37.6
0.4
30.4.
62.6
7.0
بی
85.8
10.9
From two-fluid nozzle equation.
Pfrom number of drops per cycle and flow rate.
'From use of u. S. Sieve Series screens.
..
Table 2. Conditions and Results of Puoa Sol Preparations
Run
No.
Pu(NO3 )4 Feed
g Pu g Pu/Liter
NH,OH
M
Peptization
Final Pula Sol
NO3/Pu
2.5
1.3
0.25
20
20
1.1-2.0
.1.6-1.8
2.5
2.5
0.11
2.5
2.1
3.75
5.0
0.11
ä
8
.
. 40
.
1.0
77
5.0
2.5
2.2
10
.
6.3
1.7
.
1.3
0.10
0.10
0.18
0.18
0.19
11
1.7
1.2
12-15
75
6.3
3.75
2.5
53
29
1.1-2.5
Table 3. Properties of Plutonia-Containing,
Sol-Gel Microspheres
Type of
Material
Sphere Sizes
(Microns)
Density,
% of
Theoret.
Surface
Area
Gas :
Release
1200° C
cc/8
Carbon
ppm
m2/8
96
0.020
< 10
0.144
20% PuO2-002
5% PuO2-U02
• 5% PuOz-Thoz
300-600, < 44
300-600
300-600, < 44
300-600, 50-250
96
99
0.026
< 10
0.015
0.03
Puoz
:
< 10
0.042
ORNL DWG 66-10980

REDUCTION
EXCESS HZ
0.5 M VOZ (NO3)2
0.15 M HCOOH
H-
HCOOH
300 psi
0.5 % PT ON AI,O, PELLETS
0.5
0.5 M U(IV)
URANOUS 1.0 M NO,
NITRATE
SOLUTION 10.29 Moon
0.25 M COOH
PRECIPITATION
ate
3.5 M NH OH -
AGITATION
ADJUST TO pH = 9
FILTRATION
SUPERNATE SOLUTION
- NH NO, NHOH,
AND NH, COOH
URANOUS
HYDROXIDE
PRECIPITATE
| I MOLE U/LITER
WASHING (FOUR CYCLES)
H,O, 2.5 LITER/MOLE U
(FOR 4 th WASH, ALSO
0.001 MOLE NH OH
PER MOLE U)
AGITATE FOR 5 MINUTES
FILTER TO'I MOLE U/L
WASH SOLUTION
+ NH4NO3, NH,OH,
AND NH,COOH
FILTRATION
DO NOT DRY CAKE
WASHED
~1.1 MOLE U/LITER
PRECIPITATE NO, ZU = -0.03
COOH XU = -0.03
URANIA SOL:
PEPTIZATION
HNO, 0.14 MOLE/MOLE UN AGITATE AND HEATH "O302~0.17
HCOOH, O. 1 MOLE/MOLE U-ALTO 60°C FOR 1 HOUR
Goh/U = 0.1
U(IV)/U = ~0.85
Fig. 1. Batch Agitated-Filter Preparation of Urania Sol.
-
-
-
ORNL DWG 66-10981
GEL MICROSPHERES AND LEH
NOTE:
LC INDICATES LEVEL CONTROL
2EH INDICATES 2-ETHYL-I-HEXANOL
SOL-
SOL
DISPERSER
CONDENSER
AND OFF-GAS
OVERFLOW
PHASE
SEPARATOR
2 in. MIN.
DIAMETER
COLUMN
PRODUCT
DRYER
SURGE
POT
WET
DISTILLATION EQUIPMENT
2EH
TANGENTIAL
FLOW
LN2 AND
STEAM
DRIED GEL

HO
(TO DRAIN)
WET 2EH TANK
DRY 2EH TANK
CALCINER
FURNACE
UPFLOW
CALCINED
OXIDE
MICROSPHERES
Fig. 2. Microsphere Preparation System for Coated Païticle
Development Laboratory.
ORNL DWG 66-10982

for LOUDSPEAKER
2 EH
DRIVE
FLUID
L
PLASTIC
BUCKET
VARIABLE FREQUENCY
POWER SUPPLY
VOLTS
FREQUENCY
:
C. FREE-FALL OR
"BUCKET" WITH
MULTIPLE ORIFICES

Q. TWO-FLUID NOZZLE
b. VIBRATING CAPILLARY
· Devices for Dispersion of Sol as Drops in 2-Ethyl-1-Hexano

NOW BEING
PREPARED
Fig. 4. TWO-Fluid Nozzle Disperser: 2.4 cc of Uoa Sol per Minute,
110 cc 2EH per Minute, 800-4. Droplets.
PHOTO 85526



....
- ...ogos -
----
.
.
.
---...


OT
Fig. 5. Vibrating Capillary Disperser: 9.6 cc of UO2 Sol per
Minute. 60-cos Vibration. 1370-12 Droplets..
1.
.

1.
· OANL Dwg 66-9291
!
Pu(NO3)4 FEED
1.3 LITERS 7
58 g Pu PER
LITER
WASH H20
10 LITERS
2 LITERS
PER WASH
ACID
ADDITION
2.0 LITERS
10.4 M HNO3
WATER
ADDITION
0.3-0.5
LITERS
PRECIPITATION FILTER-WASH PEPTIZATION EVAPORATION BAKING_RESUSPENSION
1.5 LITERS 5 WASHES .~2.5 LITERS TO DRYNESS ~2 HOURS AGITATE
4M NH4OH CAKE VOLUME ~30 g Pu ~150° C MAX AT A 0.12-0.32
100% EXCESS ~0.5 LITER PER LITER BREAK UP 1 UNIFORM 1 LITERS
NOz/Pu 2.5 CAKE 250° C 1-3M PUO2
CONDENSATELJ
0.1-0.3 LITERS
CONDENSATE
~2 LITERS
FILTRATË
~12 LITERS
PuO2
SOL
Fig. 6. Typical Flowsheet for Puo2 Sol Preparation.
R-31934

1.
.
.
.
•"* 34., ene
i
2
...
.
.
14!
.
!
.
.
:
.....
$
.
.
.
."
.
.
-
2
A
their webshop
.
..
7
ontslae
+
.
Hoe
1
.
.
150-200M
450-600 M
· Fig. 7. Puoz Sol-Gel Microspheres Calcined at 1200°C.

R-31935

1
s
500 u
.
.
.
.
150-200 p
450-600 u
Fig. 8. Puoa Sol-Gel Microspheres Calcined at 1200°C.

"'.
.'
."
"T.
.1: .
*
.
.
.
.
. Pr
X
...
9
ORNL DWG 66-11017-
A = OUTPUT FROM FIRST HEATUP
B: OUTPUT FROM SECOND HEATUP (BASELINE)

DIFFERENTIAL THERMOCOUPLE OUTPUT INVI

o+
Ot
500 6.00 700
100 200
400 500
700
DIFFERENTIAL THERMOCOUPLE OUTPUT (än.
AL00 500
TEMPERATURE
(°C)
TEMPERATURE
Th02 SOL-GEL MICROSPHERES AS
EXTRACTED FROM SPHERE FORMING
COLUMN
b. Tho, SOL-GEL MICROSPHERES
AFTER STEAM TREATMENT AT 152° C
FOR EIGHT HOURS
DIFFERENTIAL THERMOCOUPLE OUTPUT (UV)

100 200 300 400 500 .700
T.EMPERATURE (°C)
- GEL
RAGMENTS FREE OF ORGANIC RESIDUES
tterns
Fig. 9. Comparison of Differential Thermal Analysis
of Sol-Gel Thoz Heated at 10°C/min in Helium.
.
.ORNL-DWG 66-8222
10,000

Tho, /VAC 10° torr/25°C
10./STEAM/200°C
RESIDUAL CARBON CONTENT (ppm)
Tho/VAC 10 torr/600°c
Thoz/STEAM/200°C
Uog/STEAM/400°C
Tho,/AIR/600°C
Tho,/STEAM/400°C
10
cm
. 10
100
·
TIME (hr)
Fig. 10. Effect of Various Drying Procedures on Residual Carbon
Content of Thoria and Urania Sol-Gel Microspheres.
ORNL DING G6-12816 •
..به
ا
م i
"
بعد مدة ..
.
.. ..

1138
000000000ننننهمسرم
ا
ا . /- lo
9825 منمفومن-
و 908 ممنسجنسهممنهما
835ec منججمجمه
-
••906 م
. .
: FRACTIONAL SHRINKAGE (ALILON
TTTT
ہ مممعننه 22*°
نو 0
0
0
0
0
0
0 من نفقه 2
lo-2/
.:. الل
102
.
ا
I03
TIME (SEC)
:
Fig. 11. Isothermal Shrinkage of Sol-Gel Thoa Microspheres.
-
-
-
-
=
=
=
=
=
---
ORNL DWIG G6-12817

9900
نننننننننن
مممممممممممنم
875°C
78500 ممبممممهجمجمه
.. .
ح ممنبم
و 7
/
0
000:
FRACTIONAL SHRINKAGE (AL/L.0)
منم-منمنمهه
680°C اسے
م ممنے
. 702
Io4.
*** TIME (SEC.)
Fig. 12. Isothermal Shrinkage of Sol-Gel UO2 Microspheres.
*** Type or on
M**
*
p
young tangan
a
proposta per
a party
ORNL-DWG 66-8223R

otto
THOZ (BANNISTER)
20
uo, (DRY ARGON +4% HYDROGEN)
CRYSTALLITE GROWTH FACTOR
SURFACE AREA DECREASE FACTOR
DOO
500°C
• 600°C
o 700°C
0 800°C +
A 900°C
A 1000°C
.
10-2
10-1
. đo 3
FRACTIONAL SHRINKAGE
. .
Fig. 13. Relationship Between Changes in BET Surface Area, Crystallite
Size, and Fractional Linear Shrinkage of Sol-Gel UO 2 and ThO2.
L
IB.: 41.
4 +TENARIT TITLYN
V
ITA
E
!.
.
1
1 TK
.
.
.
.
.
...
..
.
· ORNL DWG 66– 11016

4 % HYDROGEN
DIFFERENTIAL THERMOCOUPLE OU
_400_590 690 700 89
TEMPERATURE (°C)

TI
DIFFERENTIAL THERMOCOUPLE OU
100
200
7
500
600
700
800
900
- TEMPERATURE (°C)
: A = OUTPUT FROM FIRST HEAT UP
B = OUTPUT FROM SECOND HEATUP ( BASELINE)
Fig. 14. Comparison of Differential Thermal Analysis (DTA) Patterns
of Sol-Gel UO2 Fragments Heated at 10°C/min. VO2 samples were precalcined
at 400°C for 4 hr in argon + 4% hydrogen. ,
.
END
DATE FILMED
15 / 17 / 67








V