1Phosphate glass, ²Aluminosilicate glass, ³IAEA technique, 28^th day of leaching in deionized water.

Unlike cementation being suitable for solidification of LILW with low salt content only, bituminization technology can be used for treatment of high-salt waste. But, this process, like cementation, has many disadvantages, for example, fire danger, low waste volume reduction factor (VRF), relatively high leaching of radionuclides from bituminized waste, soil contamination with nitrates, etc.

Glassy wasteforms were recognized to be more appropriate compared to cementous and bituminous wasteforms due to higher chemical durability, VRF, mechanical strength, long-term stability under hydrogeological conditions.

Radon R&D activity in field of glassy wasteforms includes development and testing of glasses for LILW immobilization, chemical analysis and determination of glass properties (viscosity, electric resistivity, density, leaching, radiation stability, etc.), and study of glass structure.

Institutional LILW contains sodium nitrate as major component (up to ~600 kg/ m³) and minor calcium-magnesium carbonate, ferrous compounds, sulfates and chlorides. To transform LILW to stable solid forms a borosilicate glass has been chosen. Boron free aluminosilicate glasses may be used as well.

Natural minerals and rocks: datolite CaBSiO₄(OH), dolomite CaMg(CO₃)₂, bentonite, and loam clay are used as glass forming additives.

Glasses obtained relate to the following forming systems:

1. Na₂O-CaO-(MgO)-Al₂O₃-SiO₂;
2. Na₂O-CaO-(MgO)-Al₂O₃-Fe₂O₃-SiO₂;
3. Na₂O-CaO-(MgO)-Al₂O₃-B₂O₃-SiO₂;
4. Na₂O-CaO-(MgO)-Al₂O₃-Fe₂O₃-B₂O₃-SiO₂.

Radioactive constituent includes ^134,137Cs, ⁹⁰Sr, ⁶⁰Co, ¹⁴⁴Ce, and traces of actinides. Waste volume activity ranges between 10⁷ and 10¹⁰ Bq/m³ on b -g -emitters, and between 10⁵ and 10⁶ Bq/m³ on a -emitters. Glasses obtained have specific activity 10⁵-10⁷ Bq/kg on b -g -emitters and 10²-10⁴ on a -emitters. It corresponds to weight concentration of radionuclides of 10^-2-10^-5 %.

The same systems are basic for conventional glasses used in industry and glass formation in these systems is studied well. A feature of waste glasses is elevated sodium content in order to reach the highest waste oxide content.

NPP wastes from RBMK and WWER type reactors were also studied. Chemical composition of RBMK waste is very close to institutional waste and it contains mainly sodium nitrate. Therefore, the above-listed systems can be applied for immobilization of RBMK waste as well. WWER waste contains both sodium nitrate and sodium tetrahydroxylborate as major components. Vitrification of this waste does not require boron containing additives. WWER waste glasses relate to systems #3 and #4.

At vitrification of sulfate- and chloride-bearing wastes phase separation problem occurs. SO₄^2- and Cl^- solubility in silicate melts is known to be approximately ~1 wt% each [5]. Excess of sulfates and chlorides results in the phase separation. Waste oxide content limitation in glass is estimated to be 5-10 wt.%. To prevent the phase separation several methods were tested. Among them:

• addition of components increasing sulfate and chloride solubility, such as lead or vanadium oxides;
• vigorous melt agitation followed by fast cooling down to upper annealing temperature to fix dispersed sulfate-chloride phase into the host borosilicate glass;
• addition of reducing agent to feed composition to decompose sulfates and remove their with off-gas.

Vitrification of sulfate-chloride-containing waste yields inhomogeneous (phase-separated) glasses or glass-composite materials consisting of dispersed sulfate-chloride radiophase distributed in matrix vitreous phase. Distribution factor for radiocesium between sulfate-chloride phase and glass is approximately 10. Study of inhomogeneous and composite materials comprises investigations of the structure of phase-separated glasses, phase analysis of glass crystalline materials and composites, and study of elements distribution among co-existing phases, determination of material properties (density, leaching, mechanical strength, radiation stability, etc.) and their comparison with properties of homogeneous and ceramic wasteforms. The main analytical methods used are atomic absorption spectroscopy, emission spectral analysis, b -g -spectrometry, a -spectrometry, infra-red spectroscopy, electron paramagnetic resonance (EPR), X-ray diffraction (XRD), scanning and transmission electron microscopy (SEM/TEM), electron microprobe analysis (EMPA), viscosimetry, electric resistivity measurements, ⁶⁰Co source accelerating facility for radiation researches. Properties of LILW glasses are summarized in Table II.

Table II. Properties of liquid LILRW glasses.

* for vitreous phase, ** if present

Solid burnable radioactive waste at Radon is burned in chamber-type incinerator with fuel heating [1]. Incinerator ash is incorporated in cement matrix. Numerous disadvantages of cementous wasteforms stimulated search of more appropriate wasteforms. The most promising wasteforms seem to be glassy or glass-crystalline materials.

Chemical composition of incinerator ash (in wt%): 2-8 Na₂O, 3-9 K₂O, 8-20 CaO, 3-7 MgO, 4-18 Al₂O₃, 3-33 FeO_n, <1-2 MnO, 1-3 Cr₂O₃, 14-38 SiO₂, <1-4 TiO₂, 2-22 P₂O₅, 2-14 ignition loss (organic residue and carbon). Specific activities on b -g -emitters and a -emitters are 10⁶-10⁷ and 10⁶-10⁸ Bq/kg, respectively. The most important alpha-emitters are ^235,238U, ²³⁹Pu, and ²⁴¹Am. Maximum concentrations in the vitrified samples are ²³⁵U - 0.02 %, ²³⁸U - 2%, ²³⁹Pu -~4× 10^-4 %, ²⁴¹Am - ~5× 10^-7 %. Traces of ²¹⁰Po and ²²⁶Ra are also present. Taking into account ash chemical composition, some flux additives for vitrification were tested (Table III). Both borosilicate and aluminosilicate glasses were designed [6]. Properties of the ash-containing materials are given in Table IV.

From XRD data source incinerator ash consists of b -whitlockite Ca₃(PO₄)₂, hydroxylapatite Ca₁₀(PO₄)₆[(OH)₂,CO₃)], calcite CaCO₃, quartz SiO₂, plagioclase, and amorphous phase. The same source ash heated to 1000 ⁰C contains b -whitlockite, hydroxylapatite, potassium aluminosilicates kalsilite KAlSiO₄ and leucite KAlSi₂O₆, and quartz. After heating to 1450 ⁰C b -whitlockite and vitreous phase were only found.

Vitrified ash melted at 1450 ⁰C with dolomite-bentonite flux contains nagelschmidtite Ca₇(PO₄)₂(SiO₄)₂ and amorphous phase. Vitrified samples are inhomogeneous. The sample vitrified with dolomite-bentonite flux was found to be more homogeneous compared to samples of vitrified unfluxed ash.

A study of waste elements distribution in the vitrified ash was also carried out. In the vitrified ash sections depleted and enriched with Ca, P, and Si and some Cr-rich inclusions were found. Matrix glass contains K, Al, Fe, and Si. It confirms XRD data showing occurrence of whitlockite or nagelschmidtite. These minerals are host phases for Sr, rare earths, Th, and U, and probably another actinides (Pu, Am). Cr-rich inclusions are possibly Cr-containing spinel that was not detected by XRD due to very small content. These may concentrate Mn, Fe, and Co.

Both glass and ceramic are considered as possible host materials for high-level and actinide wastes. Vitreous state is thermodynamically unstable. Therefore, glass may be devitrified during long-term storage. Effect of glass devitrification on radionuclide retention and safety of their immobilization is variable and not known in details. Unlike glass, natural minerals are stable for geological periods and crystalline ceramics can guarantee safe immobilization of long-lived radionuclides for long-term periods (up to hundreds of millions years).

Among crystalline host phases, various synthetic analogs of natural minerals are considered (Table V). We pay the most attention to the synthetic minerals that can be produced through melting. These are synthetic zirconolite, pyrochlore, perovskite, hollandite, murataite, britholite, sphene, brannerite, pyroxenes, iron-alumina garnets, etc., and their assemblages (Synroc).

Works on rock-type wasteforms were started at Radon in 1990. Preparation method was chosen to be high frequency melting in the cold crucible where Radon is greatly experienced. Since 1994 works are performed in cooperation with IGEM RAS.

Step 1 of work on Synroc-type materials had original purpose to prove formation of the Synroc-type ceramics at the cold crucible melting. Using melting in both resistive furnace and the cold crucible the Synroc-type material has been obtained. It contained hollandite, zirconolite, perovskite, and rutile as major phases, and powellite CaMoO₄, hibonite/loveringite CaAl₁₂O₁₉, and CAT-phase as minor phases. The material obtained was examined with optical microscopy, SEM/TEM, EMPA. No cracking and elevated leaching were found after fast neutron irradiation of the melted Synroc-C to dose 4× 10¹⁸ nvt. These investigations were carried out in cooperation with ANSTO and IGEM RAS [7,8].

The purpose of step 2 was development of Synroc formulation to immobilize PA "Mayak" HLW compositions. This work was also performed in cooperation with ANSTO and IGEM RAS.

Two Synroc formulations for immobilization of source HLW and actinide-rare earth fraction of partitioned HLW have been designed. To immobilize non-partitioned HLW conventional Synroc-C formulation containing 20 wt.% calcined HLW was used. Material was produced by hot-pressing at ANSTO and through melting at Radon. Major phases were found to be zirconolite, hollandite, rutile, and perovskite. In the hot-pressed sample minor metal alloy was also revealed.

In the melted Synroc, containing simulated partitioned HLW the same major phases were found, but minor phases were zirconia, celsian, and murataite (Ca_2.65U_0.3Ce_0.2)(Ti_7.3Mn_0.6Zr_0.4Al_0.3)O₂₀ or (in wt%): 59.8 TiO₂; 15.6 CaO; 7.0 UO₂; 5.6 ZrO₂; 4.7 MnO; 4.1 Ce₂O₃; 1.8 Al₂O₃. The latter phase concentrates uranium, especially in the central parts of grains. Uranium content in the core reached ~12 wt.% whereas uranium content in the rim was lower than 1 wt.% (~0.8 %) only. Murataite, whose content in the sample was about 1-2%, concentrated up to ~41 wt.% of total uranium in the sample whereas zirconolite (~30-40 vol.% in the sample) contained 52 wt.% U. Summary of waste elements partitioning among co-existing phases in the Synroc doped with simulated Russian HLW are given in Table VI.

Currently Synroc melting tests are carried out using bench-scale facility with high frequency generator operated at 1.76 MHz and vibrating power 60 kW. The plant is equipped with batch preparation and feeding unit, melt pouring and cast annealing unit, and off-gas system.

Step 3 was devoted to development of the cold crucible melting of zirconolite/pyrochlore-rich ceramics. We synthesized zirconolite-rich ceramics containing U⁴⁺ (Pu⁴⁺ analog) and Gd³⁺ (Pu³⁺ and Cm³⁺ analog) by means of both the cold crucible inductive melting and cold pressing+sintering. Up to 50 wt% UO₂ (Synroc-F formulation) or 30 wt% Gd₂O₃ were incorporated in synthetic zirconolite/pyrochlore.

As far as materials for plutonium immobilization are concerned, the most promising host phases for Pu immobilization are believed to be zirconolite and pyrochlore. Monoclinic zirconolite structure transforms into cubic pyrochlore structure when Pu³⁺ content exceeds 0.4 formula units in the Ca site and Pu⁴⁺ content exceeds 0.15 formula units in the Zr site [9]. The latter case corresponds to ~11 wt% ²³⁹PuO₂.

Another hosts suggested for actinide immobilization are murataite, zircon, cubic zirconia, sphene (titanite), aechynite/ euxenite (REE,Th,U,Ca)(Ti,Nb,Ta,Fe³⁺)₂O₆, brannerite UTi₂O₆, monazite, synthetic analogs of apatite and related phases (silicophosphates), NZP ceramics and analogs, borosilicate glass, phosphate glass.

Technological researches comprises

• development and testing of the cold crucible inductive melting process,
• the cold crucible design,
• development and testing of the plasma arc treatment of solid waste,
• development and testing of the plasma arc melting process,
• development of the integrated plasma treatment-inductive melting process.

The cold crucible melting process is being currently developed for both LILW vitrification [2,3] and HLW immobilization in Synroc and other ceramic and glass-ceramic materials. The plasma arc melting process is mainly designed for solidification of high-fusible solid low-level, mixed, and hazardous wastes. The integrated plasma treatment-inductive melting process is under development now for processing of unsorted solid waste whose chemical composition and morphology are widely varied. In the whole, general concept, accepted at SIA Radon, is to transform all the radioactive wastes to stable glass-like or ceramic wasteforms suitable for long-term storage and able to be transported to regional repositories. From this point of view, processes with liquid slagging are the most promising.

Radon has numerous facilities including resistive furnaces for lab-scale investigations (up to ~1 kg of batch), bench-scale facility (1-10 kg of batch) based on the cold crucible (up to 10-20 kg/h), industrial-scale facility (10-50 kg of batch) based on the cold crucible (up to 30 kg/h), bench-scale facility for plasma torch melting (up to 10-20 kg/h), industrial-scale vitrification plant (up to 75 kg/h) for production of glass blocks each of ~25-30 kg.

Future works will be directed to development and testing of appropriate wasteforms for HLW and actinide-bearing waste immobilization (Synroc, zirconolite/pyrochlore-rich ceramics, other crystalline materials), and further development of the cold crucible and plasma technologies for waste conditioning. R&D program includes:

• Design of ceramic formulations for immobilization of source (non-partitioned) and partitioned HLW. Incorporation of non-partitioned HLW in conventional Synroc-C (20 wt.% HLW), Cs/Sr fraction of partitioned HLW in Synroc-B and C (~10 wt.% HLW) or glass-ceramic based on perovskite, pollucite, and sphene, and actinide-rare earth fraction in zirconolite/pyrochlore-rich ceramics;
• Production of the materials designed by the cold crucible melting with bench-scale facility;
• Products characterization;
• Design and construction of the demonstration plant for the cold crucible melting of rock-type materials.

As far as plutonium immobilization is concerned, works are supposed to be as follows:

• Production of 0.1-1 kg samples containing long-lived ²⁴⁴Pu.
• Preparation of multiple kilogram batches of ceramics or glass with Pu surrogate and doped with small amounts of actual Pu in system where they demonstrate the same behavior.
• Study of Pu-bearing samples produced by our collaborators with XRD, TEM/SEM, IR spectroscopy, EPR, EMPA as well as optical microscopy (in cooperation with IGEM).

1. I.A. Sobolev, L.M. KHOMTCHIK, "Management of Radioactive Wastes at Centralized Sites," Moscow (1983).
2. I.A. Sobolev, S.A. Dmitriev, F.A. Lifanov, S.V. Stefanovsky, A.P. Kobelev, "Low- and Intermediate-Level Waste Vitrification: Basic Principles, Process Units and Product Characterization," WM95. HLW, LLW, Mixed Wastes and Environmental Restoration - Working Towards A Cleaner Environment. Tucson. 1995. Abstracts. P.12. CD ROM.
3. I.A. Sobolev, S.A. Dmitriev, H.A. Turlak, F.A. Lifanov, S.V. Stefanovsky, A.P. Kobelev, M.I. Ojovan, "SIA Radon Experience in Radioactive Waste Vitrification," Moscow, Enomar (1996).
4. N.P. Laverov, B.I. Omelyanenko, S.V.Yudintsev, B.S. Nikonov, I.A. Sobolev, S.V. Stefanovsky, "Mineralogy and Geochemistry of Matrices for the Immobilization of High Level Radioactive Wastes," Geology of Ore Deposits 39 (1997) 179.
5. W. Eitel, "The Physical Chemistry of the Silicates," Chicago (1954).
6. T.N. Lashtchenova, F.A. Lifanov, S.V. Stefanovsky, "Incorporation of Radon Incinerator Ash in Glass and Glass Crystalline Materials," WM97. HLW, LLW, Mixed Wastes and Environmental Restoration - Working Towards A Cleaner Environment. Tucson. 1997. CD ROM.
7. S. Stefanovsky, E.R. Vance, R.A. Day, B.D. Begg, "Melting and Ceramic Routes to Synroc," PacRim2. The 2^nd International Meeting of Pacific Rim Ceramic Societies, Abstract. P.A145 (1996).
8. O.A. Knyazev, B.S. Nikonov, B.I. Omelianenko, S.V. Stefanovsky, S.V. Yudintsev, R.A. Day, E.R. Vance, "Preparation and Characterization of Inductively-Melted Synroc," Spectrum '96. Proceedings. International Topical Meeting on Nuclear and Hazardous Waste Management, Seattle, WA, 1996. P. 2130-2137.
9. E.R. Vance, A. Jostsons, R.A. Day, C.J. Ball, B.D. Begg, P.J. Angel, "Excess Pu Disposition in Zirconolite-Rich Synroc," Mat. Res. Soc. Symp. Proc. 412 (1996) 41.

BACK