Note: Descriptions are shown in the official language in which they were submitted.
-- 1 --
CKGROUND OF INVENTION
This invention relates to a process Eor the permanent
environmentally and biologically safe storage, for extra-
ordinarily long periods of time, of highly radioactive waste
materials. More specifically this invention relates to a
process for preparinq a glass-ceramic composite product for
containment of solid radioactive wastes and to the glass-ceramic
product.
Considerable effort in recent years has been directed
at developing techniques for conditioning of liquid radio-
active wastes for disposal. In particular high-level and
intermediate-level liquid wastes from fuel reprocessing are
contaminated with fission products and long-lived lanthanides
and actinides and must be solidified for disposal. The
solidified waste products must be stable enough and disposed
of in such a way as to prevent the biologically hazardous
radionuclides from contaminating the environment over their
hazardous lifetimes. One method which has been considered
for disposing of liquid radioactive wastes involves
incorporating the wastes directly in glasses, glass-ceramics
or ceramics, then emplacing the waste products deep under-
ground in geological formations of salt, granite, shale or
basalt. In this case inactive constituents of the wastes,
which can sometimes adversely affect product durabilities,
are incorporated in the products along with the radionuclides
so that relatively large volum~s of waste must be disposed
of deep underground. Another method which has been considered
involves removing long-lived fission products, lanthanides
1~99~9L3
-- 2
and actinides from the liquid wastes using ion-exchange
materials, then converting -the contaminated ion-exchange
materials into glasses or ceramics. In this case only
relatively small volumes of ion-exchange materials contamin-
ated with long-lived con-taminants need be converted into
stable, leach resistant products for disposal deep under-
ground. The large volumes of decontaminaied liquids, now
containing only relatively short-lived radionuclides,
could be disposed oE in a less rigorous way as low-level
10 wastes.
Of the ion-exchange materials which have been
identified for use in decontaminating radioactive liquid
wastes, zeolites, titanates and calcium hydroxyapatite have
been shown to be effective at removing cesium, strontium,
some lanthanides and some actinides. Attempts have been
made to convert the zeolites and titanates to stable leach
resistant products suitable for disposal of radioactive
wastes. Such attemp-ts have included melting the zeolite with
a borosilicate glass, hot pressing calcium titanate
to form a ceramic product, and cold pressing
mixtures of radioactive contaminated zeolites (Na-
form mordenite) and sodium/ammonium titanates followed by
atmospheric sintering of the mixtures.
~1though the technology of glassmaking is well
established on an industrial scale and the procedures for
incorporating ion-exchange materials in glass are straight-
forward, glass products are not thermodynamically stable
and may be expected to dissolve slowly under the conditions
likely to be encoun-tered over very long periods of time in
a waste material repository. While great care is taken with
43
the selection of -the repository --likely a granite pluton
such as may be found in the Canadian Shield-- there can be
no guarantee that ground waters will not permeate the
repository and, at the temperatures generated within the
stored waste radioactive material r leachingand/or structural
instabilities are to be expected.
While certain phases of the ceramic products are
generally thermodynamically stable under the expected
conditions and are therefore suitable for the storage of
radioactive wastes, most ceramic products are mixtures of
many crystalline phases only some of which can be identified.
In these cases it is difficult to use thermodynamics and
geological data to assess long term stability and compat-
abllity oE ceramic products with host rock formations.
Further, the -technology of hot isostatic pressing is complex
and is not well developed on an industrial scale. The
products made by cold pressing and sintering tend to be
porous and have poor mechanical strength. For sintering to
produce an impermeable body, long sintering times and high
temperatures are usually necessary and the loss of volatile
nuclides is correspondingly high. Fabrication of a large
ceramic body is notoriously difficult, mainly because the
thermal gradients in a ]arge monolith during heating give
rise to non-uniform phase formation, differential shrinkages
etc. It is doubtful whether ceramic monoliths of good
qua]ity and of a size comparable with a standard waste
canister for a vitrified product could be made.
S MMA Y OF THE_ NVENTION
In view of all of these limitations there is a need
~9~
for alternative technology for the very long term s-torage
of radioactive wastes, and i-t is an object of the
present invention to provide a process for the production
of a glass-ceramic composite product in which the crystalline
S phase is thermodynamically stable and is compatib]e with the
host roc~. The glass~ceramic product has sufficient mechanical
strength an~ is relatively easy to make in the required
size.
Thus, by one aspect of the present invention there
is provided a process for preparing a glass-ceramic product
for storing radioactive nuclear waste materials for extra-
ordinarily long periods of time comprising:
(a) passing liquid radioactive waste materials through
a selected ion exchange medium to thereby deposit said
radioactive materials thereon and separate a relatively
low level radioactive liquid for subsequent disposal;
(b) heating said ion exchange medium bearing said
radioactive materials with sufficient glass forming
constituents being present, so as to form a meit, the non-
radioactive portion of which has a composition in the range
M2OO - 15 wt. %
M112O30 - 15 wt. %
SiO235 - 65 wt. %
Tio210 - 35 wt. %
MlO0 - 15 wt. %
CaO5 - 10 wt. %
Mlllo - 3 wt. %
Ml 250 - 3 wt. %
where M is selected from Na, K
Ml is selected from Ca, Ba~ ';r
Mll i5 selected from Al, 8, Fe, Cr
Mlll i9 selected from Zr, Sn, Zn
and MIV is selected from P, Ta, Nb;
(c) cooling saicl melt so as to form a glass; and
(d) heat treating said glass so as to crystallize
sphene crystallites in a protect;ve glassy matrix
and containing said radioactive materi~als.
By another aspect there is provided a glass-ceramic
product for storing radioactive nuclear waste materialR for
extraordinarily long periods of time, having a composition:
M2O 0 - lS wt.%
M112o3 0 - 15 wt,%
15 SiO2 35 ~ 65 wt.%
TiO2 10 - 35 wt.%
~lo 0 - 15`wt.
CaO 5 - 10 wt.%
Mlllo - 3 wt.%
20 MlV2O5 ~ 3 wt.%
where M is selected from Na, K; M is selected from Ca, Ba, Sr;
M is selected from Al, B, Fe, Cr; Mlll is selected from Zr, Sn,
æn; and MlV is selected from P, Ta, Nb; and compri~ing sphene
crystallites in a protective qlassy matrix.
By yet another aspect there is provided a cartridge for
the treatment of liquid radioactive nuclear waste materials
containing an ion exchange medium which, upon heating with
sufficient glass forming ingredients forms a glassy product
having a composition in the range
,, ~
.?4~,13
- 5a -
M2O0 - 15 wt.
M112O30 - 15 wt.
SiO235 - 65 wt.%
TiO210 - 35 wt.
M O0 - 15 wt.
CaO5 - 10 wt.~
Mlllo - 3 wt.%
MlV2O5o - 3 wt.%
where M is selected from Na, K; Ml is selected from Ca, Ba, Sr;
Mll is selected from Al, s, Fe, Cr; Mlll is sel~cted from Zr, Sn,
Zn; and ~1 is selected ~rom P, Ta. Nb.
BRIEF DESCRIPTION OF THE DRAWINGS
The invention will be described in more detail with
reference to the drawings in which:-
Figure 1 is a sketch of the phase diagram for the
3-phase system SiO2, - CaO - TiO~; and
Figure 2 is a sketch of a flow diagram of one
embodiment of the present process.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
Thermodynamic calculations and petrological-
geochemical observations have indicated that sphene,
CaTiSiO5, is an ideal host for most nuclear waste ions.
Sphene is a common accessory mineral in granites and
granodiorites, both of which are candidate rock types for
a nuclear waste repository vault, and is resistant to
chemical alteration. Analyses of mineral specimens also
indicates that the CaTiSiOS structure is capable of
accepting a wide range of solid solution substitutions
such as:
For Ca : Th, U, rare earthelements,Na, ~n, Sr, Br
For Ti : Al, Fe~+, Fe++, Mg, Nb, Ta, V, Cr, Sn,
Zr, Zn, Pb
. ~
L3
For O : O}-l, F, Cl.
Studies have also shown that sphene should be stable in the
environment of a vault in areas such as the Canadian Shield
and which is flooded with typical ground waters at a depth of
1 km or more. Thus, a sphene-based waste form is believed to
be an excellent host for wastes resulting from such opera-
tions as CANDU fuel recycling under the currently proposed
disposal conclitions.
A glass-ceramic composite produc-t has been selected
as the best compromise between ~he desirable properties of
crystalline materials and the more forgiving nature of glass
with respect to compositional variation and radiation damage
and the fact that it may be more easily prepared, via glass
making techniques, than a purely ceramic product. ~n a
preferred embodiment of the present invention, the glass-
ceramic comprises CaTiSiO5 crystallites containing at
least a substantial proportion of the longer-lived waste
ions in solid solution, surrounded by a durable glass matrix.
The CaO-TiO2-SiO5 phase diagram shows the primary
sphene crystallization field to be almost adjacent to a
large li~uid immiscibility area, and to be surrounded by
relatively low-temperature boundary lines in the range 1300-
1375C. The area of glass formation in this system is shown
in Figure 1. However, the glasses within this area are
potentially unstable, and either crystallize or exhibit
subliquidus immiscibility during cooling. In the case of
glasses that undergo phase separation during cooling, rapid
crystallization of CaTiSiO5 occurs during reheating. This
crystallization occurs uniformly throughout the bulk of the
glass (as opposed to com~encing at the surface), indicating
that the phase separation can serve to nucleate the growth
of sphene crys-tals.
Additions of ~a20 and Al203 to CaO-TiO2-SiO2 glass
formulations have been showrl -to strongly influence the rate
and extent of phase separation in the qlass. Additions of
Na20 reduce the tendency to separate, whereas A1203 additions
promote the separation. One possible explanation for this
is as follows.
The Ti ions have been shown by electron spectro-
scopic analysis to exist within the ylass network predomin-
antly as octahedrally coordinated (TiO6) groups and these
are stabilized by the close proximity of charge-balancing
Na+ or Ca2+ ions. ~lowever, the addition of A13 ions,
which enter the network as (A104) groups, creates a rival
demand for the charge-balancing cations, resulting in
destabilization of the network and an increased tendency
for phase se~aration.
Auger electron microanalysis of the separated phases
in a typical Na~O-CaO-A1203-TiO2-SiO2 glass has confirmed
the existence of an Al-rich phase and a Ti~rich phase.
Thus, by varying the Na20:A1203 ratio in these
glasses, the degree of phase separation can be controlled.
This can, in turn, influence the rate of sphene crystallization,
the possible formation of additional minor crYstalline phases,
and the glass-ceramic microstructure. The Na20:A1203 ratio
also has a significant influence on melt viscosities, although
typical melts are extremely fluid, as a result of the high
~rio2 contents.
Crystallir~.ation of a representative parent ~1ass, ,t
mole % composition 6.6 Na2O, 5.1 A12O3, 16.5 CaO, 14.8 TiO2,
57.0 SiO2, to give a sphene-based glass-ceramic, is achieved
by reheating the ~lass at a temperature between about 950C
and the melting point of the glass, and preferably in the
range 950-1050C for 0.50-1 hour. X-ray diEfraction phase
analysis and energy dispersive X-ray spectroscopic elemental
analysis have confirmed that sphene is the only crystalline
phase produced in the composition ranges under s-tudy, and
that the matrix consists of residual Na~O-CaO-A12O3-SiO2
glass. Presumahly, the extreme sluggishness of structural
reorganization in high-alumina glasses, as evidenced by their
high viscosi-ties, is responsible for the absence of further
crystallization of, for example, aluminosilicate phases.
]5 As noted above sphene may be prepared synthetically
by heating mixtures of the appropriate chemicals in a
crucible at about 1400 C for ~bout 1 hour, cooling the glass
thus produced and reheating in the range 950C - 1050C to
effect devitrification. The precise crystallization temper-
ature selected is of course dependent upon the precise
composition of the glass selected. The source of the
chemicals is largely immaterial and it has been found that
certain inorganic ion exchange media, such as zeolites,
sodium titanate and calcium hydroxyapatite, which are useful
for ion exchange processes on high level radioactive liquid
waste materials, can be heated to between about 1250C and
about 1600C and preferably at about 1400C to form a melt
which has a composition, which can be adjusted by the
addition of conventional glass making constituents as
necessary, suitable for the production of sphene therefrom.
Typically zeo]ites such as mordenite (~a-form),
Zeolon~ 900 (synthetic mordenite Na8A18Si40O96^24 El2O,
supplied by Norton Company, Akron, Ohio) are useful for
this purpose.
Thus, a cartridge containing a zeolite, such as
Zeolon~ 900 and/or soaium titanate and which contains an
insoluble source of calcium, such as wollastonite may be
used to effect an ion exchange reaction on a high-level
radioactive waste material. The high level radioactive
materials are absorbed onto the zeolite and titanate and a
relatively low level radioactivity liquid waste is discharged
for disposal in a less rigorous manner. The entire cartridge
is then heated to about 1400C to produce the desired melt,
cooled and then reheated to the range 950-1050C, preferably
about 1000C so as to effect devitrification. As the entire
ion exchange canister or cartridge is used to e~fect the
decontamination and then is treated to produce first the
glass and secondly the glass-ceramic product, the handling
of radioactive materials is minimized and simplifiedO It
will be appreciated that other sources of calcium can be used.
For example calcium hydroxide is soluble in aqueous solutions,
and could interfere witn the ion exchange reaction: it
therefore rnay be added as a solution, after the decontamina-
tion reaction is completed, and before heating to form the
melt.
One way to supply Na, Ca, as well as Si would be to
use a commercially available glass frit; e.g. a typical
soda-lime silica frit would be 14 wt.~ Na2O, 9 wt.~ CaO,
- 9a -
73 wt.~ SiO2 and A wt.% other oxides. An advantage of
this approach is that the g]ass ceramic would Melt at
lower temperatures: i.e. the glass frit would melt at
a relatively low temperature (1250C-1350C) and the
other ingredients would dissolve in the melt.
Example 1
Some tests were undertaken to determine the preferred
range of compositions for sphene-based glass-ceramics.
Products with compositions listed in ~ables 1 and 2 were
prepared by heating mixtures of p~wdered chemicals in
platinum crucibles at 1400C for 1 hour.
-- 10 --
Table 1
Some Compositions for Sphene-Based Glass Ceramics
_ _ _ . ... . .. . . _
Composition _ _Comp nent (wt._~
# Na2O CaO Ti2A123 Si2 P2o5
- .-. - --.-- -__ ____ _ ___
A 8.33 10 66 10.67 5.61 63.94 0.78
B 9.17 11.11 18.01 4.56 56.33 0.81
C 7.91 15.77 13.50 4.~5 57.59 0.78
D 8.20 13.79 14.38 4.71 58.].3 0.79
E 9.02 14.00 18.74 4.20 53.25 0.80
F 9.52 14.14 22.16 3.82 49.56 0.80
.. .. . _ ... . . _ _ . .
Table 2
The Compositions Eor the Sphene-Based Glass-
Ceramics of Table 1 Expressed as Wt.% Mordenite
(Na-form), Sodium Titanate, Bone Char and
Wollastonite
_ _ _ _ _
Composition _____ Component (Wt.~) _
# Mordenite Sodium Bone Wollastonite
(Na-form) l'itanate Char
~ ___ _
A 65.3 16.3 4 14.4
B 50.9 26.4 4 18.7
C 51.~ 20.6 4 23.5
D 54.4 21.8 4 19.8
E 48.0 28.1 4 19.9
F 43.2 32.9 4 19.9
_ __ _~ _ ___ __
The melts were coole~ rapidly to 750C by casting on an iron
plate, then r~heate~ to 1000C at the rate of 5C per minute.
The temperature was held at 1000C for 1 hour, then -the
furnace was turned of f to allow the glass-ceramics to cool.
All of the mix~s mel.e(l easily at 1400C, alth~i~yh ~ompositi.c~i
A was very viscous. 'rhe sharpest crystallization peaks as
determined by Differential Thermal Analysis were recorded
for compositi.ons B, E and F at temperatures of 940 C, 900 C
and 960 C respectively. The sole product of crystallization
as determined by X-ray Diffraction was sphene in all
compositions. ~n -the basis of the results shown in Tables
1 and 2 the range of compositions for sphene-based glass-
ceramics prepared with mordenite (Na-form) sodium titanate,
calcium hydroxyapatite and wollastonite were estimated to
be:
Component Wt.%
Mordenite (Na-form) 48% - 20%
Sodium titanate28% - 12%
Calcium hydroxyapatite 4% - 4%
Wollastonite 20% - 10%.
The following examples illustrate the production of
sphene using ion-exchange materials.
Example 2
A mixture comprising
2.2 g sodium carbonate
1.9 g alumina
12.7 g silica
3.4 g titanium dioxide
3.4 g calcium oxide
1.3 g potassium titanate
was heated in a platinum crucible at 1400C for 1 hour,
cooled rapidly to 750 C by casting on an iron plate, then
reheated to 1000 C at the rate of 5 C per minute. The
-- 12 -
temperaT~re was h~ld at L000C for 3 ~OUI-S, then Ihe
furnace was switched off to allow the glass ceramic
product to cool slowly. The product had a
composition:
Component T~t.i-
Na~O 5,4
K2O 1.2
A12O3 7.9
~io~ 53.1
TiO2 18.2
C~ 14.2.
It was deterlllined by X-ray diffraction analysis that the
only crystallization product was sphene.
Example 3
15 A powdered mixture comprising:
55.0 q Zeolon(~) 900 (sodium form)
22.0 g Sodium titanate
20.0 g ~ollastonite (CaSiO3)
was heated in a ceramic mould at 1400C for 2 hours, cooled
rapidly to 750C by removing the crucible from the furnace,
then reheated to 1000C at 5C per minute, soaked at 1000C
for 3 hours and furnace cooled. The glass-ceramic product
had a composition:
Component ~t '~
Na2O 8.3
2 3
SiO2 ~5.8
rrio2 18.3
CaO 9.9
and X-ray diffractioIl analysis showed sphene as the sc,i~
crystalline produc-t.
Example 4
. . .
A powdered mix-ture comprising
65.0 g Zeolon('~! 900 (Na-form)
23.0 ~ Sodium titanate
17.2 g Calcium hydroxide
was heated in a ceramic crucible at 1400 C for 1 hour,
cooled rapidly to 750 C by casting on an iron plate, then
reheated to 1000C at 5C per minute, soaked at 1000C for
3 hours and furnace cooled. The glass-ceramic product had
a composition:
C~ponent Wt.
Na2O8.9
A1238.3
SiO248.7
TiO220.0
CaO14.1
and X-ray diffraction analysis showed sphene as the sole
crystalline product.
Example 5
_,.
powdered mixture comprising
67.4 g Zeolon~ 900 (K-form)
~9.1 g Potassium titanate
23.2 g Calcium carbonate
was heated in a ceramic crucible at 1400 C for 1 hour,
cooled rapidly to 750C by removing the crucible from the
urnace, then rehea`ed to 1000 C at 5C per minute, soaked
at 1000C for 3 hours and then furnace cooled. The glass
ceramic prodllct ha~l a composition:
C~mpor.ent Wt. %
K2O 14.2
A12O~ 6.9
Si2 40 7
rrio2 26.4
CaO 11.8
and X-ray ciiffraction analysis showed sphene as the sole
crystalline product.
Example 6
A powdered mi~ture containing lant.anum
was heated in a ceramic crucible at 1400C for 1 hour,
cooled rapidly to 750C by removing the crucible from the
furnace, then reheated to 1000C at 5C per minute, soaked
at 1000C for 1 hour and then furnace cooled. The glass-
ceramic prvduct had a composition:
Component Wt
.
Na2O 5.84
CaO 11.61
2 3
TiO2 16.89
SiO2 48.92
l-la2o39.31
x-ray dif~raction analysis showed sphene as the sole
25 crystalline product. Scanning Auger microscopy has been
used to distlnguish all of the lanthanum and has verified
that it is contained substantially in the crystal phase.
Lanthanum in the glass was below the detection limit.
It will be appreciated that particular rererence
- 15 -
has been made herein to the Na2O ~ Al.2O3 ~ SiO2 - Tio2 -
CaO system as there is a high probability of Na2O in the
waste materials anc'. an a]uminosili.cate glass has
particularly advantageous properties. However, the present
invention is not limited -to this system as it is possible
to replace Na wi-th K cations, at least some of the Al
with Fe and Cr (natural sphenes are known to contain Fe2O3,
for example), at least some of the Ti with Zr, Sn, Zn, and
the P with T.~ and Nb.
While particular reference has been made herein to
inorganic ion exchange media, it will be appreciated that
the present invention is not limited thereto, as organic
ion exchange resins such as Duolite~ ARC 359 (Diamond Sham-
rock) and Amberlite~ IRN-15Q or IRN 154 ~Rohm & Haas) may
also be used in the production of glass-ceramic products.
Amberlite~ IRN-150, IRN-154 and IRN-300 are typically used
to purify the moderator and primary heat transport systems
of CANDU~ nuclear reactors. In exchange resins IRN-150, 154,
300 are mixed bed resins comprising both cation exchange
resins and anion exchange resins. However, cation exchange
resin alone could be operative.
Example 7
A mixture comprising:
30 g. Amberlite~ IRN-154 (Rohm & Haas)
8.4 g. Na2CO3
8.0 g. A12O3
28-Q cJ. SiO2
17.1 g. ~rio2
12.0 g. CaO
- 16 -
was heated in a ceramic crucible at 700C for 1 hour to
incinerate the resin. The mixture of ash and chemicals
was then heated at 1400C for 1 hour, cooled rapidly to
750 C by casting on an iron plate, then reheated to 1050 C
at 5C per minute, soaked at 1050C for 1 hour and furnace
cooled. The glass-ceramic product had a composition:
Co~nent~t.~
_ _ _
Na2O 7.0
A123 11.5
S12 40 0
Tio2 24.4
CaO 17.1
X-ray diffraction analysis showed sphene as the sole
crystalline product.
