In addition to the issue of treatment of surplus weapons grade materials, increased interest is being shown in the immobilization of special categories of waste arising from the pyrochemical reprocessing of Pu metal for weapons use. These differ from the wastes generated during the reprocessing of spent fuel as they can contain high concentrations of actinides together with substantial quantities of halides, particularly chlorides, as illustrated in Table 25.7, which gives the compositions of simulated salt wastes under investiga­tion at the UK’s Atomic Weapons Establishment (AWE).

Wastes containing large quantities of chloride are not amenable to immo­bilization in borosilicate glass because of the very low solubility of chlorides in this type of glass. Similarly, Synroc-type ceramics cannot be employed either due to low halide solubility.

Two approaches can be taken when dealing with this type of waste. One is to remove the halides prior to immobilization of the non-halide constitu­ents employing, for example, borosilicate or phosphate glass; the second is to accommodate the halides chemically in a suitable host. Halides can be removed by a number of methods, but one disadvantage of this route is that secondary waste is produced which must also be dealt with. One example is reaction of the waste with ammonium dihydrogen phosphate to yield ammonium chloride and water as by-products, together with a phosphate glass (Donze et al., 2000):

2NH4H2PO4 + MCl2 ^ MO. P2O5(glass) + 2NH4ClT + 2H2OT

Another example is use of lead silicate glass to yield lead chloride as a volatile by-product, the chloride reacting with PbO in the glass, and the resulting oxide dissolving in the glass (Forsberg et al., 1997):

3PbO + 2PuCl3 ^ 3PbCl2 T + Pu2O2

Iron phosphate-based glasses have also been suggested for immobilizing chloride wastes, but in the UK AWE’s experience, the bulk of the chloride is not incorporated chemically but is evolved during waste form processing, again generating a secondary waste. Calcium aluminosilicate-based glasses have also been suggested (Siwadamrongpong et al. , 2004) and have been shown to be partially effective in incorporating chloride constituents, immo­bilizing up to 17.5 mol% calcium chloride, for example (Schofield et al., 2008). Unfortunately, regardless of the amount of chloride in the initial batch, up to 30% of the chloride is evolved during the melting process, making this method no more attractive than methods suggested for remov­ing chloride prior to immobilization.

Alternatively, halides can be chemically incorporated into a number of ceramic hosts including chlorapatite, Ca5(PO4)3Cl and spodiosite, Ca2PO4Cl, with the actinides being incorporated into the substituted whitlockite-type phase (Donald et al., 2007), which is one of the methods being investigated at AWE. An alternative method involves occluding the halide species into a zeolite and heating above 800°C to form the sodalite mineral phase Na8(AlSiO4)6Cl2 (Lewis et al., 1993; Metcalfe and Donald, 2004). This method has been adopted by the Argonne National Laboratory (ANL) for immobilizing pyrochemical wastes arising from reprocessing of experimen­tal fast breeder reactor fuel, where salt-loaded zeolite is mixed with glass and converted into a monolith by either hot-isostatic pressing or pressure­less sintering. The phase assemblage produced by both processes is essen­tially the same, consisting of about 70% sodalite, 25% binder glass and 5% halite and oxide inclusions (Lewis et al., 2010). A similar method was inves­tigated by AWE for immobilizing weapons-related pyrochemical wastes but was rejected in favour of the phosphate route. In the case of the calcium phosphate immobilization route, waste powder may be reacted with calcium phosphate to yield a mixture of chlorapatite and spodiosite: for example:

PuCl3 + 8Ca3 (PO4)2 ^ 2Cas(PO4)зCl + 4Ca2PO4Cl + Ca6Pu(PO4)б

The resulting powder will subsequently be encapsulated in a sodium aluminium phosphate or similar glass to yield a monolithic product.

Fluidized bed steam reforming has also been suggested for treating halide-containing wastes (e. g., Jantzen, 2003). The product from this process is a highly durable waste (Jantzen, 2006) consisting of a number of feld — spathoid phases having cage stuctures (e. g., nephelines and sodalite), which contain the halides.