A very important thermophysical property to be considered for an engineering material, like nuclear fuel, is its melting point. The onset of melting at the cen­terline of the fuel rod has been widely accepted as an upper limit to the allowable thermal rating of nuclear fuel elements [11, 12]. The melting point must be taken into account when considering a new fuel, as it limits the power that can be extracted from the fuel element. The knowledge of the melting point is also important in the fabrication of chemically homogeneous pellets like thoria-urania since ThO2 (3,663 K) and UO2 (3,100 K) have high melting points and relatively low diffusion coefficients at normal sintering temperatures [13].

As a pursuit for the better fuel, it is crucial to understand the underlying transport phenomena due to electrons and lattice vibrations in actinide systems.

According to the Lindemann criterion, solids with large Debye frequencies have high melting points [14]. This is typically found in insulators where atomic bonds are strong due to lack of free electrons. Thorium contains no occupied 5f states while uranium has two unpaired 5f valence electrons, and therefore uranium and thorium possess very different electronic and chemical properties. Since melting point is an important property, it is worth considering from the standpoint of the bonding present in actinide elements [15]. The highest melting points for the actinide metals are for Th and Pa metals. The effect of f-orbitals on the melting point is maximized with Np and Pu; both have very low melting points, which are believed to be a reflection of f-orbital repulsion [16]. Uranium has multiple oxidation states (3+ through 6+) which allow UO2 to be easily oxidized to U3O8 or UO3, by incorporating interstitial O atoms. In contrast, thorium only exhibits one oxidation state (4+) and hence cannot be oxidized beyond ThO2 [9, 17]. Among the actinide oxides, only ThO2 is a white insulating solid and the other AnO2 solids are all dark and opaque and poorly conducting. While in UO2 the 5f electrons, which occupy an energy band from 1.37 to 1.50 eV cause a strong visible light absorption resulting in a dark gray color of this oxide. The consequent absence of low electrons in the valence band is the cause of the high transparency of stoichiometric thoria and the low spectral emissivity in the visible range at room temperature [17].

Thorium dioxide exists up to its melting point as a single cubic phase with the fluorite crystal structure, isomorphous, and completely miscible with UO2. Unlike UO2, ThO2 does not dissolve oxygen to a measurable extent. Therefore, it is stable at high temperature in oxidizing atmosphere. On prolonged heating to 1,800-1,900 °C in vacuum, it blackens with loss of oxygen, although the loss is insignificant to be reflected in lattice parameter or in chemical analysis. On reheating in air to 1,200-1,300 °C, the white color is restored.

The melting points of the nuclear fuels are shown to be influenced by the following factors: stoichiometry and composition, irradiation dose, impurities and their contents.

2.1 ThO2

The melting point of ThO2 was experimentally measured or estimated by several authors [18-26]. Their results are summarized in Table 1. As it can be seen from the Table, the reported values vary from 3,323 to 3,803 K. Peterson and Curtis [26], in their compilation of data on thorium-based ceramics, arrived at two dif­ferent values, e. g., 3,573 ± 100 K from the work of Lambertson et al. [21] on ThO2-UO2 system and 3,663 K from the work of Benz [22] on Th-ThO2 system. Lambertson et al. [21] first estimated the melting point of ThO2 to be between 3,558 and 3,828 K and subsequently arrived at an intermediate value of 3,623 K by extrapolating the melting point data of (Th, U)O2 compositions corresponding to zero UO2 content. They further refined their data by introducing some corrections

for the liquidus/solidus curve to effect a curvature correction for the pure ThO2 end to that of pure UO2 end of the temperature—composition diagram. Their final recommended data was 3,575 K, which is in good agreement with the data 3,543 K recommended by Christensen [27] from his experimentally measured melting point data on ThO2-UO2 system and subsequent extrapolation to zero UO2 content. Rand [25], however, disagrees with the curvature corrections made by others on the thoria or urania rich side of the temperature composition curve. He justified that the curvature need not be same at both the terminal compositions and the difference could be due to the loss of ‘O’ from UO2 in urania-rich side, which is different for thoria-rich side. He recommended a value of 3,643 ± 30 K. Belle and Berman [12] used 3,640 K as the melting point of ThO2, recommended by Rand [25] in his work on ThO2. Ronchi and Hiernaut [24] had recently measured the melting temperature of ThO2 (both stoichiometric and hypostoichiometric) material experimentally by heating a spherical sample by four symmetrically spaced pulsed Nd YAG laser and observing the cooling/heating curve with time. For stoichiometric ThO2, the measured melting point was found to be

3.651 ± 17 K. The data of Ronchi and Hiernaut [24] reasonably agrees with the data generated by Benz [22] (3,660 ± 100 K) and is also close to that recom­mended by Rand [25] (3,643 ± 30 K). All these values are markedly different from those of Lambertson et al. [21]. It is also well understood that the curvature difference at the uranium — and thorium-rich side of the temperature versus com­position diagram is quite justifiable and was attributed to the loss of oxygen. Hence, the recommended melting temperature of ThO2 should be taken as

3.651 ± 17 K, and is in fairly good agreement with majority of the previous studies. The value measured at the Institute for Transuranium Elements (ITU) (3,640 ± 20) K, is very close to the value reported by Rand [25].

Measurements of the cooling curves of molten ThO2 and ThOi.98 reveal that the stoichiometric compound melts congruently at 3,651 K, while the hypostoichio — metric oxide displays a liquidus at 3,628 K and a solidus transition at 3,450 K. Ronchi et al. [24] conducted pulse-heating experiments on thoria and showed that this compound exhibits a premelting transition at 3,090 K whose features are analogous to those observed in other ionic compounds having fluorite type structures. A class of diatomic compounds which crystallize in the face-centered cubic fluorite lattice (space group Fm3m), and whose component atoms have very different mobilities, exhibit at a temperature corresponding to about 80 % of the absolute melting temperature, a premelting transition. The discovery of this transition in UO2 [12] gave rise to a number of investigations aimed at defining its nature and possible effects on the high temperature properties of this technologi­cally important material. In ThO2-x, the dependence of the transition temperature, Tc on stoichiometry was found to be weaker than hypostoichiometric urania and exhibits an opposite trend, with Tc decreasing with x.

To conclude, the melting temperature of ThO2 recommended from this assessment is 3,651 ± 17 K and is in fairly good agreement with majority data available in the literature.