No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

The problem of high-temperature volatilisation of material from alkali borosilicate glasses, commonly used for the long-term storage of high-level nuclear waste, has been investigated using four mixed-alkali borosilicate glass systems. Each system is based on the mixed alkali borosilicate base-glass (MW) used in the UK and USA, initially being doped with one of three simulated waste oxides in varying quantities; aluminium oxide (0.95 mol%), lanthanum oxide (1.68 mol%) or magnesium oxide (2.55 or 10.2 mol%), before increasing quantities of caesium o~de were added (0-9.66 mol%) [1]. A number of experimental techniques have been employed in this study, including various thermal analysis measurements, nuclear magnetic resonance (MAS NMR) and Raman spectroscopy. From boron-ll and silicon-29 MAS NMR measurements, the fraction of tetrahedral boron (N4) and silicon Q3 and Q4 units (with respective three- and four-bridging oxygen) were resolved and compared against theory [2] and previous experimental measurements [3-6]. Silicon Q3 fractions resolved from NMR measurements [1] were compared against values obtained from quantitative Raman spectroscopy [7]. This approach is shown to be successful in resolving silicon Q3 units and therefore would be particularly useful in studies where the presence of large (> 1 mol%) quantities of paramagnetic species in glasses makes NMR difficult. Further detail resolved in the lIB MAS NMR spectra of these glasses also enabled the fraction of reedmergnerite and danburite medium~~~' order structures to be calculated [8]. The volatilisation of alkali borate and alkali silicate material, identified using EDX measurements, is shown not only to increase as a function of the fraction of Q3 units but is also affected by the presence of these medium-range order structures, where increasing quantities of danburite units leads to a greater degree of volatility [4,8]. Finally, a number of 170-enriched mixed-oxide alkali borosilicate systems have been manufactured, using an 17O-enriched silicon oxide precursor obtained through an established hydrolysis reaction of 20% oxygen-I7 enriched water with silicon tetrachloride [9]. Subsequent thermal analysis and NMR measurements have shown the oxygen-I7 environments of these borosilicate systems to vary significantly based on the presence of caesium, aluminium or lanthanum in quantities realistic to the modified mixed-alkali borosilicate glasses in the remainder of this study.

The phase diagrams of Zn7Sb2012 doped with Ni, Cr and Co were determined. In Zn7-xMxSb2012, Ni to substitutes up to x = 4; Co forms a full solid solution range. Cr substitutes up to x = 0.35 in Zn7-4xSb2-2xCr6xOI2. Substituting each metal ion for zinc has the effect of decreasing the a -+ Pphase transition temperature from 1225 °C in undoped Zn7Sb2012. Cr has the most drastic effect; only a small'amount of Cr, ~0.075, is required to stabilise the a-phase at all temperatures. For,no Cr-containing composition of was the phase pure p-polymorph observed, although a small compositional range yielded a mixture of a and Ppolymorphs. In both Ni and Co doped solid-solution series, no Ppolymorph was observed above x = 2. At high cobalt concentrations, a small amount of Co oxidises from C02+ to C03+, implied by the presence of the antimony-rich phase CoSb206. As temperature -is increased, the amount of C03+ decreases; at ~1200 DC, a single phase XRD pattern was observed, corresponding to the phase containing only the reduced form of cobalt, Co2+. The ternary subsolidus phase diagrams ZnO - Sb20s with one of NiO, CoO or Cr203 have been determined. Each consists of a number of solid solution regions, most notably in the Co case, in which five solid solution regions are observed. The rest of the diagrams consist ofa number oftwo- and three-phase compatibility triangles. Up to four Ni ions can substitute for Zn in Zn7Sb2012, each occupying octahedral sites according to the formula (Zn3)tet[Ni4Sb2]octOI2, determined by Rietveld refinement using neutron diffraction data. No evidence was found to suggest the presence of tetrahedral Ni. Cobalt was found to occupy both octahedral and tetrahedral sites. Co has a small preference for octahedral geometries, but the entropy contribution at high temperatures means some tetrahedral occupation becomes favourable. The preferential octahedral site occupancy of Co lead to a negative deviation from linearity in a Vegard's law plot oflattice parameter, a vs composition. Impedance spectroscopy measurements showed that doping Zn7Sb207 with any of the ions studied has the effect of increasing the conductivity by a small amount, although none of the samples are good semiconductors. The density of the pellets of each composition was low, leading to constriction resistance in many of the samples. Cu is observed to substitute for Zn up to x = 4. At low Cu concentrations the p-phase exists, in which Cu is thought to replace Zn on octahedral sites. At 900°C, a new phase, Zn3CU4Sb2012, forms. By use of Visser's algorithm with XRD data, this phase has a monoclinic unit cell, space group Cc and lattice parameters a = 21.01777 A, b = 8.7802 A, c = 5.58475 Aand P= 112.5912°. The structure was solved using direct methods followed by combined Rietveld refinement using XRD and time-of-flight neutron diffraction data. The detailed formula is the same as that of spinel, with Cu occupying the octahedral sites and Zn occupying the tetrahedral sites, according to the formula (Zn3)tet[CU4Sb2]octOI2. The octahedral linkage can be described as a cationdeficient rock salt structure, similar to that of the p-Zn7Sb2012. The tetrahedra share edges, forming infinitely-linked 'columns' parallel to the c-axis. Many ofthe sites are fractionally occupied, leading to more sites containing cations than in spinel.

No description available.

No description available.