A Thesis on the Constitution of Alloys of Thorium with Certain Transition Metals. , by J.R. Thomson Submitted for the Degree Of

A thesis on

The constitution of alloys of thorium

with certain transition metals. ,

J.R. Thomson

submitted for the degree of Doctor of Philosophy in the University of London.

Department of Metallurgy, Imperial College of Science August, 1963. and Technology, Prince Consort Road, London, S.W.7. 1.

SUMMARY

The behaviour of the transition series of elements has long been of interest and the position of thorium in group 4A of the periodic table near to the beginning of a possible 'actinide' series of elements has made a knowledge of its properties and alloying behaviour of especial interest from the theoretical point of view. In the present work, the physical properties and electron theory of the transition metals and their alloys have been reviewed. Factors affecting the formation of intermetallic compounds have also been reviewed and their applications to some specific types of crystal structure discussed.

The phase diagrams of six binary alloy systems of thorium with the metals ruthenium, rhodium, palladium, osmium, iridium and platinum have been investigated in the temperature range 1000°C - 1500°C, mainly by x-ray and metallographic methods. Solid solubility of thorium in palladium has been observed and intermetallic compounds occur in each system. The crystal structures of 15 of these compounds, embracing 7 structure types, have been determined while the structures of 5 other compounds have been confirmed. In several cases additional information about the latter compounds has been obtained. 2. In general, the results are in agreement with alloy theory although the solubility of thorium in palladium would not be predicted by current theories. There are several points of similarity between the present systems but the constitution of Th-Pd alloys is often different from that of the other five systems. The thorium-rich eutectic compositions and temperatures show a general correlation with the melting points of the component phases similar to that reported for eutectics formed by magnesium, the rare earths, uranium and plutonium. Available results on the structures of the intermetallic compounds have been discussed in relation to other compounds with the same structures and the phase diagrams have been compared with the available information on several other alloy systems of the group 8 metals. The structures of compounds containing less than 50% group 8 element tend to be characteristic of the other element and size factors tend to be important in determining which structures will occur. With more than 50% group 8 element, size factors are still important but the structures can often tolerate a much wider range of radius ratio. 3. The Constitution of Alloys of Thorium with certain Transition Metals.

Summary. List of contents. List of tables. List of figures. 1.Introduction. 2.Literature review. 2.1. The transition elements. 2.1.1. Physical properties. 2.1.2. The electron theory of the transition metals. 2.1.2.1. Collective electron theories. 2.1.2.2. Localised wave-function approach. 2.2. Factors affecting the formation of intermetallic compounds. 2.2.1. General. 2.2.2. Th7Fe3 (D 102) structure.

2.2.3. CuA12 (C 16) structure. 2.2.4. CrB (Bf) structure. 2.2.5. Ni2In (B 82) structure. 2.2.6. Laves phases MgZn2 (C 14), MgCu2 (C 15) and MgNi2 (C 36). 2.2.7. (L 12) structure. Cu3Au 2.2.8. TiNi3 (D024) structure. 2.2.9. CaZn5 (D 2d) structure. 4. 2.3. Alloys of the transition metals. 2.3.1. Solid solutions. 2.3.2. Alloys of intermediate composition. 2.3.3. Alloys of the platinum metals. 2.3.4. Alloys of thorium. 2.3.5. Other work on the systems under investigation. 3.Experimental methods. 3.1. Alloy preparation. 3.2. Heat treatment. 3.3. Thermal analysis. 3.4. Metallographic preparation. 3.5. Hardness. 3.6. X-ray techniques. 4. Results. 4.1. The thorium-ruthenium system. 4.1.1. The phase diagram. 4.1.2. The crystal structure of Th7Ru3. 4.1.3. The crystal structure of ThRu. 4.1.4. The crystal structure of ThRu2. 4.2. The thorium-rhodium system. 4.2.1. The phase diagram. 4.2.2. The crystal structure of 0-ThRh2. 4.2.3. The crystal structure of ThRh3. 4.3. The thorium-palladium system. 4.3.1. The phase diagram. 4.3.1.1. Alloys of 0-75% palladium. 4.3.1.2. Alloys of 75-100% palladium. 5.

4.3.2. The crystal structure of Th2Pd. 4.3.3. The crystal structure of Th3Pd5. 4.3.4. The crystal structure of ThPd3. 4.3.5. The crystal structure of ThPd4. 4.4. The thorium-osmium system. 4.4.1. The phase diagram. 4.5. The thorium-iridium system. 4.5.1. The phase diagram. 4.5.2. The crystal structure of ThIr5. 4.6. The thorium-platinum system. 4.6.1. The phase diagram.

5. Discussion. 5.1. The phase diagrams. 5.1.1. Accuracy of the results. 5.1.2. Points of similarity between the systems. 5.1.3. The thorium-rich eutectic temperatures. 5.1.4. Melting points of the compounds. 5.1.5. Age-hardening behaviour and the lattice parameter of thorium. 5.2. The intermetallic compounds. 5.2.1. Compounds of the type Th7X3. 5.2.2. Compounds of the type Th2X. 5.2.3. ThX compounds with the CrB (Bf) structure. 5.2.4. The compounds Th3Pd5 and Th3Pt5 5.2.5. The compound p-ThRh2. 5.2.6. The Laves phases (AB2). 6. 5.2.7. Compounds of the type ThX3. 5.2.7.1. Compounds with the Cu3Au (I, 12) struAure. 5.2.7.2. Compounds with the TiNi3 (D024) structure. 5.2.8. Compounds of the type ThX5. 5.2.9. The types of crystal structure observed. 5.3. Comparison with expectations. 5.4. Comparison with other alloy systems. 5.4.1. Alloys of thorium with iron, cobalt and nickel. 5.4.2. Alloys of titanium, zirconium and hafnium with group 8 metals. 5.4.3. Alloys of uranium and plutonium with group 8 metals. 5.4.4. Alloys of lanthanum and cerium with group 8 metals. 5.5. The behaviour of palladium in intermetallic systems. 5.5.1. General. 5.5.2. Thorium-palladium alloys.

6.Suggestions for further work.

7.Conclusions. 7.1. The thorium-ruthenium system. 7.2. The thorium-rhodium system. 7.3. The thorium-palladium system. 7.4. The thorium-osmium system. 7.5. The thorium-iridium system. 7.6. The thorium-platinum system. 7.7. Crystal structures of the compounds. 7.8. General. 7

Appendix 1. Acknowledgements. References. Tables. Appendix 2. Figures. 8.

LIST OF TABLES

Table 1. Classification of atoms with respect to their electron transfer properties. 2. Comparison of observed and calculated sin2 0 values and line intensities for Th7Ru3. 3. Interatomic distances in Th7X compounds. 3 4. X-ray data for compounds. Th7x3 5. Comparison of observed and calculated sin26 values for Th7Rh3, Th70s3' Th71r and 3 Th Pt . 7 3 6. Comparison of observed and calculated sin26 values and line intensities for ThRu. 7. Structural data for ThX compounds having the ) structure. CrB(Bf 8. Comparison of observed and calculated sin26 values and line intensities for ThRh. 9. Comparison of observed and calculated sin26 values and line intensities for ThIr. 10. Comparison of observed and calculated sin26 values and line intensities for ThPt. 11. Interatomic distances in ThX compounds. 12. Comparison of observed and calculated sin26 values and line intensities for ThRu2. 13. Comparison of observed and calculated sin26

values and line intensities for ThOs2 and ThIr2. 9.

Table 14. Interatomic distances in ThX2 compounds. 15. Comparison of observed and calculated sin26 values and line intensities for p-ThRh2. 16. Comparison of observed and calculated sin26 values and line intensities for cubic ThRh3. 17. Comparison of observed and calculated sin; values and line intensities for tetragonal ThRh3. 18. Phases identified in Pd-Th alloys. 19. Hardness of Pd-Th alloys after homogenising for 4 days at 1000°C. 20. Comparison of observed and calculated sin26

values and line intensities for Th2Pd. 21. Comparison of observed and calculated sin26 values and line intensities for Th Pd 3 5 22. Comparison of observed and calculated sin26 values for Th3Pt5. 23. Comparison of observed and calculated sin26 values and line intensities for ThPd 3 (75% p alladium). 24. Comparison of observed and calculated sin26

values and line intensities for ThPd4 (80% palladium). 10.

Table 25. Comparison of singe values for alloys of 66-83% iridium. 26. Comparison of observed and calculated singe values and line intensities for ThIr5. 27. Summary of closest interatomic distances in A°. 11.

LIST OF FIGURES

Figure 1. The thorium-ruthenium system. 2. The thorium-rhodium system. 3. The thorium-palladium system. 4. The thorium-osmium system. 5. The thorium-iridium system. 6. The thorium-platinium system. 7. Eutectic temperature vs melting point of the compound. 8. Eutectic composition vs melting point of the compound. 9. Eutectic temperature vs mean electron concentration. 10. Interatomic distances between unlike atoms in

Th2X compounds. 12.

1. Introduction In recent years the development of new branches of technology has called for many new materials; this has led to commercial interest in a number of metals which occur in the transition series of the periodic classification and also in the so-called heavy metals thorium, uranium and plutonium. The increased availability of these metals in relatively pure form and the desirability of improving their properties has led to a study of many of their alloy systems.

Thorium is of interest in the nuclear energy field since it has a naturally occurring radioisotope Th232. This cannot sustain a chain reaction by itself, but in a reactor 233 in the presence of slow neutrons, thorium can breed U by neutron capture and subsequent decay. 233 0 233 0 233

In the pure state thorium is very soft, 40 VPN, and as the chemical or alloy state of an atom does not affect its nuclear properties, there is considerable interest in the possibility of strengthening thorium by alloying. This thesis reports the results of investigations on the 6 binary alloy systems of thorium with the transition metals ruthenium, rhodium, palladium, osmium, iridium and platinum. The thermal neutron absorption cross-sections of the 13. platinum metals are high (Ru 2.461 Rh 150, Pd 8.0, Os 14.7, Ir 440 and Pt 8.1 barns/atom) which precludes their use as intentional alloying additions to thorium in a reactor but ruthenium, rhodium and palladium are formed in small amounts as fission products on bombardment of U233 by thermal neutrons. There is therefore some practi- cal interest in alloys of those metals with thorium although the primary interest in each system has been to assist alloy theory.

When the work was started, nothing was known about the alloys of thorium with the platinum metals except for the structure of two intermetallic compounds ThOs2 and ThIr2. Consideration of the atonic sizes of the elements suggested that mutual solid solution between thorium and the platinum metals might be restricted and the differences in the electrochemical properties indicated that there might be a tendency to form stable intermetallic compounds. While the work was in progress, information on the structures of several additional compounds was published and showed considerable agreement with some of the present results as discussed later.

2. Literature review 2.1. The transition elements. In the periodic classification,of the elements, the 14. transition elements are characterised by the fact that in the free atoms, the energy levels of the n(s) and n-1(d) electrons are vary close together and are both less than the n(p) levels. As a result of this, in the third and following elements after the rare gas in the first and second long periods, the (d) shell is progressively filled up to its full complement of 10 electrons. In the first long period, the transition process leads to the building 2 6 up of the (3s) (3p) (3d)10 group of 18 electrons which is the maximun number of electrons which can occupy the 3-quantum shell. In the second long period the transition process builds up a group of 18 4-quantum electrons, a 2 further expansion of the 4-quantum shell to 2x4 = 32 electrons takes place in the third long period by the filling of the (4f)14 sub-group in the rare-earth series of elements which lie between lanthanum and hafnium.

These simple considerations apply to the free atoms of the elements; when the atoms approach to the distances found in the solid crystal and in the liquid near the melting point, the experimental evidence of soft x-ray spectroscopy shows that the energy levels are broadened into bands which may overlap. The electrons are then in hybrid (spd) states, the proportions of the three components varying with the position in the period, with an increase in the amount of d-function in the later members 15. of a period.

The position of the elements of atomic number 89 and upwards (the actinides) in the periodic table has been of considerable interest since these elements were first discovered. Some members of the series (actinium, protoactinium and the elements americium and upwards) are still available only in very small amounts and although much of their chemistry is known, present knowledge of their metallurgy is fragmentary. A detailed review of (1) these elements was given by Katz and Seaborg and (2) Waldron has surveyed the results of later research. In their chemical and physical properties the elements from thorium to neptunium show similarities to transition elements while the later actinides show, in such properties as oxidation states, a marked resemblance to the trend of similar properties in the lanthanide elements.

A generalization which accounts for many aspects of the behaviour of the heavy elements was put forward by Friedel(3). He suggested that there is a progressive transition from pure 'd' orbital behaviour in the early members through 'd-f' hybridized or combined orbital behaviour to 'f' orbital or lanthanide behaviour in americium and later elements. In support of this hypo- thesis of a gradual transition, is the reduced stability of the b.c.c. phase favoured by transition metals and the 16. occurrence of many allotropic modifications in the intermediate zone. When the difference in binding energies for the 5f and 6d shells is small it becomes comparable with the difference in binding energies of different compounds and atomic arrangements. Hence the distribution of the electrons between energy levels may vary for a given element from one allotropic phase to another or from one compound to another.

2.1.1. Physical properties. In the crystal structures assumed by the transition metals, polymorphism is common but it is only in group 4A (titanium, zirconium, hafnium and thoriun) that it persists throughout the group. The b.c.c. form occurs most frequently and is often the high temperature form of a polymorphic metal. Apart from some complicated structures, the only other forms shown are the typically metallic f.c.c. and c.p.h. structures, the latter having an axial ratio rather less than the ideal close-packed ratio of 1.633. Considering the crystal structures in relation to the periodic table, from group 3A to group 6A the tendency is towards less close-packed structures, i.e. from f.c.c. and c.p.h. to b.c.c., after that the trend reverses and the structures return to the close-packed form of copper, silver and gold. 17.

Properties such as compressibility, coefficient of thermal expansion, atomic diameter, heat of sublimation and melting point are generally considered to give an indication of the cohesion of the atoms in pure metals. Plots of these properties against the atomic numbers of the elements show a general trend towards a maximum in the cohesion halfway through a period although the three periods differ in detail. The maximum is less well defined in the first long period than in the other two and its magnitude increases from the first to the third period.

Ferromagnetism is shown by iron, cobalt and nickel and some chromium and manganese compounds are also ferromagnetic. Ferromagnetism is not shown by the transition metals of the later periods although some of the rare-earth metals, e.g. gadolinium,are ferromagnetic at low temperatures and many of the others are strongly paramagnetic. The majority of the transition metals are weakly paramagnetic.

The magnetic properties of solid solution alloys afford a method of investigating the electronic structure of the solute atoms. From a consideration of the paramagnetic susceptibility of palladium and that of its alloys with the noble metals, hydrogen and nickel, Wohlfarth(4) concluded that there are 0.6 holes/atom remaining in the 18. d-band of palladium and that these holes are filled up by the outer s electrons of the added metals or hydrogen. Later investigations of the effects of alloying elements on the magnetic susceptibility of palladium have been made by Hoare et al(5), Wucher(6), Bates and Leach(7) and Gersten- berg(8).

2.1.2. The electron theory of the transition metals. The qualitative picture of the electron configuration of the transition metals is fairly clear. Up to the stage where breaks in the physical properties occur , there seems to be general agreement that all of the outer electrons are involved in cohesion. After this stage, the two methods of approach, the collective electron theories and the hypotheses put forward by Pauling based on localised wave-functions, start to differ in their interpretations.

2.1.2.1. Collective electron theories. In these theories the breaks in the physical properties are regarded as due to a marked increase in the repul- sive characteristics of the wave-functions. Use is made of Bloch functions in which the wave-function for each electron extends through the whole lattice; the functions consider the electrons singly and in treating any one electron, the fields due to the remainder are regarded as smoothed out. 19.

Depending on the approximations used, there are three methods of approaching the problem:-

a) the nearly free electron approximation b) the tight binding approximation c) the cellular method of Wigner and Seitz.

In principle, the cellular method is considered to be the best but in practice, with more than one electron per atom, there are great difficulties in obtaining a self- consistent field and it is seldom that good agreement is obtained between calculated and observed results. Attempts have been made to derive the d- and s- band widths for certain of the transition metals but the results differ according to the method used. Hume-Rothery and Coles(9) point out that the width of the d-band almost certainly increases with decrease in the atomic number.

2.1.2.2. The localised wave-function approach On this approach, the breaks in the sequences of physical properties are assumed to result from the entry of some of the outer electrons into atomic or non-bonding d-orbitals, these being distinct from the remaining d-orbitals which are regarded as forming hybrid spd bonding orbitals whose resonance gives rise to the metallic bond. This is not a 'theoretical' approach in the true sense of the word; it is a scheme involving assumptions chosen to 20.

fit the experimentally deteriined properties of the elements considered. The chief assumption is that the wave-function for each electron is localised around an individual atom or pair of atoms although exchange of electrons can occur allowing movement through the lattice as in the collective electron theories.

For the transition metals, the number of bonding electrons per atom is taken to be equal to the group number up to group 5A and then fitted to the magnetic data for the later elements. The number of bonding electrons and the relative energy levels of the s, p and d bands change with the total number of electrons. This results in a variation in the contribution to the bonding from each function and therefore to a change in the character of the bonding hybrid across the periods. This is the basis of the scheme (10) proposed by Altmann, Coulson and Hume-Rothery to explain the crystal structures of the transition elements.

2.2. Factors affecting the formation of intermediate compounds In this section it is proposed first to mention briefly some of the factors which control the occurrence of intermetallic compounds and then to give a more detailed account of current ideas on some of the structures assumed by compounds in the six systems under investigation. 21.

2.2.1. General. (11) Pauling has extended the ideas of electron transfer to intermetallic compounds by dividing the elements in the periodic table into four classes; stable, hypoelectronic, hyperelectronic and buffer atoms as shown in Table 1. Hypoelectronic atoms have more bond orbitals available than valency electrons and can increase their valency by accepting electrons. Hyperelectronic atoms have more valency electrons than bond orbitals and can increase their valency by giving up one electron of a pair occupying a bond orbital, thus leaving a valency electron in the orbital. Buffer atoms can give up or accept an electron without changing their valency; however, chromium, molybdenum and tungsten are buffer atoms with respect to the addition of an electron only. Buffer atoms have partially filled d-states and can either give up a non- bonding d-electron or introduce an electron into an incomplete d-shell without a change in valency. Carbon and silicon have stable valencies of four and either the addition or the removal of an electron causes a decrease in their valency.

Waber and Gschneider have applied these principles of electron transfer to the lanthanide(12) and actinide(13) elements (thorium, uranium and plutonium). They observed that, if boron, aluminium, beryllium and magnesium were 22. considered to be hyperelectronic for these elements, hypoelectronic atoms do not form compounds with the rare earths and actinides; the hyperelectronic and stable atoms do form compounds while the buffer atoms show a gradual transition from non-compound forwers to compound formers.

(14) Pauling has also pointed out that interatomic distances in metallic crystals are usually of the same order as those to be expected from resonating covalent or one-electron bonds. He therefore considered metallic bonding as an extension of the resonance process as applied to covalent compounds and he related interatomic distances to bond numbers by the empirical relation Dn = D1 - 0.60 log10 n where n is the bond number, Dn is the interatomic distance corresponding to this number and D1 is the single bond interatomic distance. He has pointed out that the appli- cation of this idea is complicated by the fact that some compounds have structures which make it possible that some of the interatomic distances represent bonds in tension and others represent bonds in compression.

(15) In a recent symposium, Raynor made a comprehen- sive review of the types of structures which occur in alloy systems in relation to the bond mechanisms involved. In 23. (16) the same symposium Kubaschewski and Sloman reviewed the relations between certain physical properties and the bond mechanism of intermetallic compounds; particular attention was given to electron density distributions, electrical conductivity, magnetic susceptibility and the energy changes associated with compound formation.

Several factors have been shown to affect the type of compound formed between two metals; these include the ratio of the atomic radii of the component atoms, the valencies of the components, the degree of chemical interaction between them and the increase of coordination on formation from the component metals.

The types of bonding mechanism which have to be considered are a) polar or ionic bonds, b) covalent or homopolar bonds, c) metallic bonds and d) van der Waal's forces. The crystal structure adopted by a particular intermetallic phase is rarely due to a single factor and it is only when one factor greatly outweighs the others that a clear connection between bond mechanism and structure can be established. In most cases the type of structure dictated by one factor may be distorted to satisfy the conditions due to another factor, in fact the forces holding a structure together are frequently mix- tures of mechanisms. This is shown by the fact that the 214-. reasons for the occurrence of two isostructural phases need not necessarily be the same, e.g. CsC1 and CuZn have the same structure although their behaviour and properties are dissimilar.

Kubaschewski(17) has looked at co-ordination in intermetallic compounds on a thermodynamic approach and suggested that the stability of metallic multicomponent phases comes mainly from a decrease in the heat content due to an increase of co-ordination on formation from the component metals. Structures discussed on this basis included the Laves phases, CaZn5, Oyu and TiNi3.

(18) It was pointed out by Kasper that a number of the relatively complex structures, particularly of the transition metals, may be regarded as determined by the geometrical requirements for sphere packing. This idea (19) was developed by Frank and Kasper who considered in detail the co-ordination polyhedra which occur most frequently. They also examined the ways in which these shells may be joined together and they applied these principles to an analysis and classification of represen- tative structures. Among the lattices which they discussed are the Laves phases, CaZn5 and the several related

A2B17 structures and the 6-phaseor 0-uranium structure. 25.

(20) Kripyakevich also reviewed the co-ordination around the individual atoms in metals and intermetallic compounds and found that one can isolate a class of structures in which some of the atoms have a co-ordination of 12 and others have a co-ordination greater than 12. He described the co-ordination polyhedra which have more than 12 vertices and analysed the relationships between them. This review covered many of the structures analysed by Frank and Kasper, including those mentioned above.

Schubert and his school considered the geometry of intermetallic compounds in a somewhat different light. They have assumed that while electrons must be free to move through the lattice, there is a tendency for the electron-gas to be concentrated in the spaces between the atoms, i.e. to have a structure or "positional correlation". They further assumed that electrons in the outer shells, i.e. the normal valency electrons, and those in the outer sub-group of the next shells may all take part in the bonding. While such assumptions may be open to criticism the results of reviews of many isostructural series of compounds show that many correlations do exist on this basis. Amongst the structures which they have reviewed are the (21) (22) Laves phases and some ordered structures 26.

2.2.2. Th7Fe3 (D 102) structure Although this structure was not analysed by Frank and Kasper, it is one in which large (M) atoms and small (X) atoms are packed together with a high co-ordination around the M atoms. The radius ratio of large to small atoms is in the range 1.30-1.47. There are three sets of non-equivalent positions for the larger atoms, each of which has 15 neighbours, the detailed types of neighbour differing for each set of positions. The X atoms have 9M neighbours and there are no close contacts between X atoms.

(23) Florio discussed the structure of Th7Fe3 in relation to Brillouin zone theory; one difficulty here is to estimate the number of bonding electrons in the unit cell since there would be 56 using Hume-Rothery valences and 90.7 using Pauling's values for the valences. From the observed interatomic distances, valences of 4.03,

4.44 and 3.42 were obtained for Th Th11 and Th111 atoms respectively and 4.23 for the transition metal using Pauling's method of calculating the bond numbers.

Compounds known to crystallize with this structure (24) (25) include Th 7Fe 3(23) ' Th 7Co 3' Th„5i7- - 3' Cer?Ti 3 ' Re7B3 ' Rh (26) 7B3 and Ru7 B 3 • 27.

(0 16) structure. 2.2.3. CuAl2 Laves(27) surveyed 24 compounds of this type (AB2) and found that the ratio of the atomic radius of the larger (B) atoms to the smaller (A) atoms lay within the range 1.09-1.54 with an average value of 1.24. Geometrical considerations, assuming that neither the A nor the B atoms should be able to "rattle" in the structure, give limiting radius ratios of 1.08-1.50 in good agreement with the observed limits. Since Laves' survey was made, several additional compounds have been found to have this structure and their radius ratios all lie within the above limits.

A characteristic of the structure is that each B atom has one B neighbour at a distance considerably closer than the atomic diameter but Laves considers that the lattice constants are a consequence of A-B interaction and that the observed A-A and B-B distances follow from the lattice parameters. In a discussion of interatomic distances in the XTh2 compounds AlTh2' AgTh2 and AuTh2 where 0 the atomic radius of each X atom is approximately 1.43 A (28) Murray pointed out that the X-Th distances decrease with increase in the electronegativity of X (i.e. the X-Th distance decreases with increased X-Th interaction), and that from the geometry of the lattice there should be a corresponding increase in the axial ratio as observed. 28.

The total co-ordination around the larger atoms in this structure is increased to 15 while that around the smaller atons is 10. Compounds crystallizing with (27) NiZr (29) this structure include RhSn2 RhPb2' PdPb2, - 2 ' (30) (31) NiHf2' InTh2 and ZnTh2

2.2.4. CrB (Bf) structure.

Pearson(32) stated that by choosing the appro- priate axes and origins, the CrB(Bf) and CaSi(Bc) structures become identical and in a detailed survey of lattice parameters, atomic parameters and interatomic distances Kripyakevich(33) has confirmed this and has pointed out that the TlI (yellow modification) structure is also identical. Compounds of this type therefore include CrB, CaSi, ThAl, ThCo, HfAll PuNil CeNil DyGa and ZrNi.

In a recent survey of borides and silicides of the transition metals, Aronsson(34) has pointed out that the occurrence of the CrB structure depends to a rather

large extent on a favourable radius ratio y- = 1.43. In x practice it has been observed in alloys whose components have radius ratios within the limits 1.11-1.61.

In this lattice, the co-ordination around the larger Cr(M) atoms is 17 while that around the smaller 29.

B(X) atoms is 11, the X atoms forming zig-zag chains through the structure so that each X atom has two X neighbours. In the borides and silicides and in ThAl this X-X distance is considerably shorter than the atomic diameter while in the other compounds listed above the X-X distances are greater than the atomic diameter. Edshammar(35) has suggested that the short B-B distances in CrB might be due to B-B bonds of a covalent type. He has pointed out that the difference between the Al-Al dis- h 0 tances in HfAl(2.86A) and ThAl(2.46A) might be due to the rTh r relative sizes of the atoms 1.26 and Cr = 1.41 Al rHf being substantially greater than-Al = 1.11. However, the analogous crystal structure of ThCo does not support this 0 assumption since the Co-Co distances are 2.77A in spite of the atomic radius of cobalt being less than that of aluminium.

In most compounds with this structure, there is a considerable contraction in the M-X distance and Kirkpatrick(29) suggests that high packing efficiency plus appreciable M-X interaction are probably responsible for their stability. 30.

2.2.5. Ni2In (B82) structure The Ni2In structure is closely related to the B81 NiAs lattice by filling up the trigonal voids in the latter structure, the completely filled up version having the composition A2B. If the trigonal positions are incompletely filled, it is possible for the structure to exist over a range of composition or at the non- stoichiometric composition. Raynor(15) has pointed out that in the process of filling up, the axial ratio decreases and approaches a limiting value of approximately 1.22 and also that compounds with these structures tend to occur when the A partners have incomplete d-shells and the ions of the B partners have some degree of deformability. The metallic character is strongly developed in compounds which have a low axial ratio and a tendency to A2B composition. An additional factor which may help to stabilize the NiAs structure at certain axial ratios is the antiferromagnetic alignment of spins of the d-shell electrons as discussed by Zener(36)

In the Ni2In structure, the co-ordination around the A atoms in the trigonal positions and around the B atoms depends upon the relative sizes of the atoms and can be 17 or 11 according to the radius ratio of the component atoms. The structure has been observed in 31. systems with a wide range of radius ratios, e.g. rZr in Zr 1 --- . 1.11 while in Pd Pb 7rurPd = 0.81. Other rAl 3 2 compounds with this structure include Rh3Sn2, Ti2Sn and Pd2Tl.

2.2.6. Laves phases MgZn2(C14), MgCu2(C15) and MgNi2(C36). The "Laves phases" are a family of close-packed compounds of approximate formula AB2; from geometrical considerations, the ideal radius ratio of the components r A should equal 1.225, but the structure is observed when rB this ratio varies between 1.05 and 1.68. The name "Laves phase" covers three closely related types(37).

C14 Hexagonal, MgZno-type, with packing ' sequence of layers PQPQ C15 Cubic, MgCu2-type ti II 17 PQRPQR C36 Hexagonal, MgNi2-type II II PQPRPQPR

In all three structures there is an increase in co-ordination on formation, that of the A and B atoms being 16 and 12 respectively. The co-ordination numbers are independent of the structure but the particular arrangement of A and B atoms around the B atoms differs in each type although all contain tetrahedra of B atoms.

Most of the original work on the relationships between these structures was carried out by Laves and 32. his collaborators(38739) with magnesium as the A element. Progressive replacement of copper in MgCu2 and of zinc in MgZn2 by elements of lower and higher valency showed that the sequence of structures:- C36--.71015 -->C14 generally occurred as the electron:atom ratio increased from: <0.92-->1.08-1.76-41.81-1.922.0 provided that the transitional elements were assumed to have valencies close to 0.

Berry and Raynor(40) listed many C14 and C15 Laves compounds and showed that, in many cases, their formation depended on a size-factor relationship and the C15 structure was more likely to occur as the ratio of the atomic radii deviated from the ideal value. The electron:atom ratios became more important when the atomic size-factors were favourable as in the magnesium alloys studied by Laves et al. A further indication that the geometrical interpretation is incomplete is given by the observation that nickel does not foie Laves phases with titanium, zirconium, niobium or tantalum even though the radius ratios are favourable(41) while uranium forms Laves phases with manganese, iron, cobalt, nickel, rhenium, osmium and iridium but not with platinum, palladium, ruthenium or rhodium(42) 33•

Raynor(15) has suggested that the phases previously classified as Laves phases are really of two types:- (a) The components have either no d-electrons or a completed d-shell; in these compounds, geometrical considerations are paramount and the interatomic distances are close to those in the pure components, e.g. KNa2, KBi2, CaLi2. (b) Compounds containing one or more transition metals where interactions between uncompleted d-shells may superimpose further conditions on the purely geometrical ones and in which interatomic distances may differ appreciably from those in the pure components. In these compounds the geometrical conditions seem to be necessary rather than sufficient and specific interactions between components may occur in a large number of cases.

(43) Recently Dwight reviewed the radius ratios and interatomic distances in 43 Laves phases and found that the radius ratio had little influence on the structure type. He suggested that the positions of the partner elements in the periodic table was the dominant factor which controlled the formation of any one of the three Laves phases, but the occurrence of the Laves phases as a group is dependent upon the ability of the 34. partner atoms to undergo the necessary contraction or expansion required to approach the effective radius ratio of 1.23. 'Dwight also pointed out that the elements palladium and rhodium are much more reluctant to form Laves phases than are their counterparts cobalt, nickel, iridium and platinum and he suggested that this may be due to an inability of palladium and rhodium atoms to undergo the necessary contraction.

0/) Bardos et al have suggested that the average number of electrons per atom outside of the closed shells of the component atoms may be an important factor in determining whether or not a Laves phase can occur at all in a given system. On the assumption that silicon can behave as an acceptor of electrons, they investigated whether or not certain Laves phases which do not occur in 12 binary nickel and cobalt systems could be stabil- ised by the addition of silicon. In each case a MgZn2- type phase was observed.

Other investigators(43,45, 6) have examined some quasi-binary systems between Laves phases in attempts to establish some of the limiting conditions governing the formation of this type of compound. The results cannot be accounted for by simple considerations of size-factor or electron:atom ratio. 35.

2.2.7. Cu Au (L1 ) structure. 3 2 This AB3 structure is an ordered arrangement of the simple f.c.c. lattice in which the repeat sequence of the layers is PQPQ; it occurs in compounds which melt congruently, e.g. TiRh3, in those formed peritectically, e.g. UA13, and it may be derived from a random solid solution by ordering in the solid state, e.g.0u3Au.

(47) Dwight has considered the lattice parameters of 12 compounds with this structure formed between transition metals and has shown that in those compounds the observed interatomic distances between unlike atoms are less than those calculated for C.N. 12 and that this difference increases with the difference in the atomic diameters of the component atoms. When the difference in the atomic diameters is small, the co-ordination around both types of atom is 12 but when the A atoms are appreciably larger than the B atoms, the larger atoms have 6 additional like neighbours at "a" giving them a total co-ordination of 18.

This type of ordering can occur at compositions other than the stoichiometric AB as evidenced by the 3 fact that superlattice lines of this type have been observed on x-ray films of Cu-Au alloys over a range of composition. In some systems the ordering is only 36. partial and may lead to a tetragonal distortion as in AlPu where the axial ratio g is greater than one, to 3 a larger cubic unit cell as in CuPt3 or to a tetragonal distortion with the axial ratio less than one as in E-CuTi3. Faulting has been observed in some systems, (48) e.g. in PdCu3, on quenching from 300-470C and it appears that in those systems the ideal superlattice cell is capable of existence only in a fairly narrow range of electron:atom ratio.

2.2.8. TiNi3 (D024) structure. This structure illustrates another way in which ordering can be achieved in a close-packed lattice when there is a size difference between the atoms; here, the repeat sequence of the layers is PURPUR. As far as is known the structure has not been observed to form as a result of ordering of a disordered solid solution. In this lattice the larger atoms have a co-ordination of 18 while that around the smaller atoms is 12.

In a review of close-packed ordered structures in binary AB3 alloys of transition elements, Dwight and Beck(47) included six compounds of this type. They showed that, as for the Cu Au structure, the observed 3 interatomic distances between unlike atoms were smaller than those calculated for C.N. 12 and that this 37. difference increased with the difference in the atomic diameters of the component atoms. Although not as common as the Cu Au lattice, this structure has been 3 observed in several binary compounds between elements of group 4A or uranium and nickel, palladium or platinum, e.g. TiNi3, ZrPt3, UPd3.

Dwight(49) has recently examined the axial ratio of the binary TiNi (1113 ) phases and of ternary alloys 3 3 in which some of the larger (A) atoms have been replaced by atoms which are neighbouring in the periodic table and which therefore have one additional or one fewer 'bonding' electrons. He has shown that the axial ratio is dependent on the relative sizes of A and B atoms and also on the position of the partner elements in the periodic table.

In the gold-rich corner of the Au-Cd-In system, (48) Schubert showed that the type of close packed struc- (A ture which occurs 1, DO24' A3 or DO19) and the axial ratio of the hexagonal phases depends on the valence electron concentration assuming gold, cadmium and indium to contribute 1, 2 and 3 electrons respectively. The

DO24 structure was observed at an electron:atom ratio of 1.22-1.25. Owing to uncertainties with regard to the number of 'valence' electrons in the metals of group 8 38.

it is not possible to interpret the structure of the binary compounds of this type in terms of electron:atom ratio.

2.2.9. CaZn5___CD2d) structure. This is another hexagonal two-layer structure in which close-packing of large and small atoms is achieved; the best packing occurs when the axial ratio a""0.817. (43) Dwight surveyed the radius ratios and interatomic distances in 27 compounds with this structure. He showed that the occurrence of CaZn (AB ) compounds is closely 5 5 dependent on the ratio of the Goldschmidt radii and that the compounds exist between the ratios of 1.29-1.61. The co-ordination around the large atom is increased to 20 while that around the smaller atom is 12. The structure can be described as derived from the hexagonal close-packed lattice of zinc by the substitution of a large calcium atom for three zinc atoms so that it contains chains of zinc atoms joined in a tetrahedral arrangement, in close similarity to the MgZn2(C14) structure but packed differently. ThFe , ThCo and ThNi have 5 5 5 this structure while hexagonal Th2Ni17, monoclinic Th2Fel7 and and tetragonal ThMn Th2Co17 12 can be derived from this lattice by the systematic replacement of particular thorium atoms by pairs of transition element atoms(23) 39.

Dwight(43) found that in the AB compounds, the 5 B element copper exhibits a lower contraction than does nickel or platinum and that palladium does not occur as a B element.

2.3. Alloys of the transition metals. Without a general theory of the transition elements themselves there is, as yet, no real theory of their alloys, even on a qualitative basis, although examination of alloy systems can sometimes help to explain some of the properties of the pure metals.

2.3.1. Solid solutions. Solid solutions in transition metals are subject to the size-factor principle put forward by Hume-Rothery et al °) to account for the alloying behaviour of silver and copper; the formation of wide ranges of solid solution is also prevented where the electrochemical factor is large. In general, solid solutions in the c.p.h. modification of transition metals are more restricted than those in the f.c.c. or b.c.c.forms. For alloys of transition metals with one another, there is a general tendency for primary solid solution to be controlled by the size-factor and by the difference between the numbers of the groups to which the solvent 40. and solute belong. Solid solutions of transition metals in other metals are generally limited.

2.3.2. Alloys of intermediate composition. This is taken to include binary alloys lying outside the limits of random solid solution and within the composition range 25-75 % of one constituent. In general, solid solution alloys of transition metals with one another tend to form superlattices to an extent which is greater than would be expected from differences in atomic diameters or electrochemical properties, e.g. superlattices have been observed in Pd-Ir, Pt-Ir and Fe-Co. The ordered structures tend to become more stable as the size-factors increase and in many systems, ordered compounds of the Cu3Au type form directly from the melt, e.g. TiRh3 and TiIr . Several other types of structure 3 which are essentially close-packed ordered arrangements, have been observed and these include the TiA1 ZrA1 3' 3' TiNi and SnNi structures. Other structures which occur 3 3 frequently include (3-W, Laves phases and ‘--phases.

Consideration of the compounds of nickel and zinc in relation to some isostructural compounds led to the suggestion that nickel might exert a zero valency in Ni5Zn21 and a valency of one in NiZn. Home-Rothery and Coles(9) have suggested that such behaviour might be 41. quite general for transition metals.

Consideration of the crystal structures of binary compounds formed between aluminium and the elements chromium to nickel led Raynor(37) to postulate a system of negative valencies for these elements corresponding to the Pauling d-band vacancies. Hume- Rothery and Coles(9) criticized this approach but Massalski(51) has pointed out that they really objected to the adoption of negative valencies as large as those proposed and that this does not affect the general proposal for the existance of electron transfer in these compounds.

2.3.3. Alloying behaviour of the Platinum metals. The platinum metals all have high melting points ranging from 1773°C for platinum to 3045±30°C for osmiu02), (Douglass and Adkins give 301000±1000(53)). No polymorphism is known and ruthenium and osmium have the c.p.h. structure while rhodium, palladium, iridium and platinum are all f.c.c.; their atomic radii are similar, varying 0 between 1.33 for ruthenium and 1.38 A for platinum. The physical, mechanical and chemical properties of the platinum metals have been surveyed by Powell(54) while their alloying behaviour with each other, with manganese, iron, cobalt, nickel and with the metals of groups IVA, 42.

VA and VIA have been reviewed recently by Raub(55)

While many binary systems of the platinum metals have not yet been investigated fully, the available information shows that complete series of solid solutions occur at high temperatures between the f.c.c. platinum metals themselves and also between them and the f.c.c. modifications of iron, nickel and cobalt; in many cases these solutions order at lower temperatures. Ruthenium and osmium also form a complete series of solid solutions. Appreciable solid solution in all the platinum metals is shown by chromium, molybdenum, and tungsten and in palladium and platinum by copper, silver, gold and cadmium, while uranium and titanium dissolve to more than 10 atomic % in palladium. Liquid immiscibility has been reported to occur in alloys of ruthenium(56) rhodium(57) and iridium(58) with silver and gold. In addition, a considerable number of stable intermetallic compounds are known.

2.3.4. The alloying behaviour of Thorium. Thorium melts at 1755±10°C and undergoes a phase transformation at approximately 1360°C being f.c.c. below this temperature and b.c.c. above. The basic factors affecting solid solution and compound formation in alloys of thorium have been described by Murray(59) 43.

On the simple Hume-Rothery size-factor approach, the 0 relatively high atomic diameter of thorium (3.59A) excludes many elements from forming extensive substitutional solid solutions and its electropositive character leads to a strong tendency to form intermetallic compounds. On this basis, apart from zirconium and hafnium in group IVA, only the rare-earths, scandium, yttrium, indium and thallium would be expected to form appreciable substitutional solid solutions with thorium.

The information available on the phase diagrams prior to 1958 has been summarized in a compilation by (42) Rough and Bauer . Complete solid solution in 0-thorium is shown only by zirconium and in a(-thorium only by cerium. Plutonium dissolves to a maximum of 47% in 04-thorium and thorium to 5% in 8-plutonium whilst hafnium is soluble to approximately 17% in 0-thorium and 7% inot-thorium. More recently, Evans has shown that -thorium forms a complete (60) range of solid solutions with lanthanum 1 that yttrium dissolves in o.-thorium to 50% and thorium dissolves to (61) 30% in yttrium , and that thorium and ytterbium are (62) (30) virtually completely immiscible . Murray has reported solubility of 9.2% indium in thorium. No information is available on the systems Th-Sc or Th-Tl.

Simple eutectic systems with restricted solid 44. solution and no intermetallic compounds occur in alloys of thorium with chromium, molybdenum, niobium, tantalum, titanium, tungsten and vanadium whilst in the thorium- uranium system there is a range of liquid immiscibility. Intermetallic compounds have been reported in many other binary systems.

The effect of several elements on the .A/(3 transformation temperature of thorium has now been established. (42) Rough and Bauer mention that the temperature is increased rapidly by. additions of carbon and decreased by zirconium, hafnium, tantalum and niobium, although it was uncertain whether the decrease caused by tantalum and niobium was a solid solution or a gettering effect. More recently, Bannister and Thomson(63) have confirmed the above results by thermal analysis. They also showed that indium and silicon additions did not change the transformation temperature significantly while aluminium, cerium and uranium lowered it. The rate of lowering caused by cerium was significantly smaller than that due to uranium, hafnium, zirconium or niobium.

2.3.5. Other work on the systems under investigation. When this work was started, nothing was known about 45. the alloys of thorium with the platinum metals except for the structure of two intermetallic compounds, ThOs2 and (64) and as far as was known no other work was con- ThIr2 templated. While the present work was in progress, information on the structures of several additional compounds (43) (65) and ThPd was published elsewhere; that on ThRia2 3 appeared after the present author had determined the structures of those compounds but the present results had (66) not been published. Independent infol.mation on T Th 0s and Th71r3(67) appeared after the structures of 7 3 those compounds had been published by the present author (68) in the open literature. The structures of Th2Pd , (43) ThRh (65) and ThIr were given elsewhere before the 3 5 present author had made any attempt to determine the structure of those compounds. 46.

3. Experimental Methods. The experimental methods used at high temperatures in the present investigation were essentially those described by Murray and Williamson(69)

3.1. Alloy preparation. All alloys were prepared by arc-melting on a water- cooled hearth using a reduced pressure of approximately a quarter of an atmosphere of zirconium-gettered B.0.0. 99.999% argon. Except for the 20 gm. specimens used for thermal analysis, all alloys were prepared as 1 gm. buttons which were subsequently broken or slit into four pieces for heat-treatment, metallographic and x-ray examination. Details of the sources and analyses of the starting materials are given in Appendix 1; the platinum metals were pre-melted prior to being weighed for alloying. For alloy preparation, each button was turned over and re-melted at least three times and the argon atmosphere was re-gettered between each melting. Weight losses on alloying were generally less than 1% of the total material employed and if this loss were exceeded in a critical alloy, a repeat specimen was prepared. 47.

3.2. Heat treatment. For heat treatment, solid specimens and powders were enclosed in tantalum capsules and sealed in silica tubes under a vacuum of better than 5x10-6 mm. mercury; the silica was outgassed at a red heat under the vacuum before the samples were sealed off. Annealing up to 1500°C was carried out in a platinum tube furnace; the temperatures were controlled by a proportionating thermocouple controller using a Pt/Pt 13% Rh thermocouple and cold-junction temperatures were controlled by a Sunvic cold-junction device. Independent Pt/Pt 13% Rh thermocouples were used for setting up the furnace and for recording the variation of temperature during the anneals. Inaccuracies in the annealing temperatures are due to a) thermocouple limitations, and b) furnace fluctuations, and are estimated to vary between ±3°C at 1000°C and ±12°C at 1500°C. Above 1375°C the silica tended to collapse around the tantalum capsules but no reaction took place and in one hour devitrification did •not extend more than half-way through the silica.

Where specimens were homogenized at 1000°C, annealing times of at least 100 hours were employed but for the solidus studies, times of one hour were found to be sufficient. Powders for x-ray examination were strain- 48. relieved for one hour at the temperature at which the parent specimen had been annealed. After annealing was completed, capsules containing powders were removed from the furnace and allowed to cool in the laboratory while those containing solid specimens were plunged into a bucket of cold water but were not broken under water since this tended to cause fragmentation of brittle samples.

For age hardening studies, specimens were wrapped in tantalum sheet and placed in a small silica tube with a chromel/alumel thermocouple strapped to the outside of the tube. This assembly was suspended in a vertical, continuously evacuated silica tube, the vacuum being maintained at better than 3x10-5 mm. mercury during ageing. An external furnace was used to heat the samples and was interchangeable with a bucket of cold water for cooling. The furnace temperature was maintained at 600±2°C by a resistance thermometer controller using a platinum resistance.

3.3. Thermal analysis. The equipment used for differential thermal analysis has already been described in detail(69,70) Two 20 gm. arc-melted buttons of the same nominal 49. composition were broken up and the pieces placed in a thoria crucible containing a re-entrant thermocouple sheath. This was positioned above a molybdenum comparator block in a molybdenum coil furnace and the whole assembly, protected by suitable molybdenum radiation shields, was housed in an evacuated, water- cooled stainless steel belljar. When the furnace temperature was 1200°C a vacuum of 5x10-5mm. mercury was indicated by an ionisation gauge positioned just below the belljar. The voltage of the power supplied to the whole equipment was stabilized and the furnace was controlled by a motor driven variable transformer; a heating and cooling rate of approximately 3°C/minute was employed. Pt/Pt 13% Rh thermocouples were placed in the crucible and in the 'molybdenum block and the temperature of the specimen and the differential temperature between thespecimen and the reference block were recorded continuously; manual potentiometer checks were taken at approximately half hour intervals.

3.4. Metallographic preparation. Specimens were mounted in a cold setting metal- lurgical mounting plastic and ground on wet silicon carbide papers down to 600 grade. This was generally followed by polishing on nylon suede cloth impregnated 50. with 6/then 1/1diamond paste but surfaces for age- hardening studies were prepared by polishing with "Silvo" metal polish on selvyt cloth. In general, phase relationships were revealed in the "as-polished" condition by examination in ordinary reflected light or under polarised light but the structure of high-palladium alloys was clarified by etching in concentrated HNO for times up to 3 10 seconds depending on the composition of the specimen.

3.5. Hardness. Hardness measurements were carried out with a standard Vickers diamond pyramid hardness tester. Loads of 5 kgm. were used for ageing studies and 10 kgm. for Th-Rh and Th-Pt alloys while 5, 10 or 20 kgm. were employed for palladium-rich alloys depending on the hardness of the particular specimen. Generally, the hardness figure quoted is the mean from five impressions but with a few alloys severe cracking occurred and it was not possible to make as many as five impressions. For ageing studies, the hardness was determined on each specimen after h111 2,4,81 16,32,64 and 100 hours at 600°C; the specimen surfaces were ground and polished after each anneal to eliminate any hardening effects caused by surface contamination. Microscopic examination was carried out after 8,32 and 100 hours annealing. 51.

3.6. X-ray techniques. Powders for x-ray investigations were obtained by two methods: a) crushing, and b) filing and annealing.

a) Brittle alloys were crushed with an agate pestle and mortar, generally the material remained bright in the laboratory atmosphere and could be crushed in air but a few alloys tarnished rapidly and these were crushed under CC14. Since sharp diffraction patterns were obtained from those specimens in the as-crushed condition, they were not annealed after crushing.

b) Powders for terminal solid solution lattice parameter determinations were obtained by filing under

0014 and any iron particles were removed with a magnet. The powders were then annealed in vacuo at the temperature at which the parent material had been annealed as described previously.

Rod specimens were prepared by sealing the powders in Lindemann glass capillaries.

For phase identification generally, a 9 cm. Unlearn powder camera was used with nickel filtered copper radiation. Where patterns were complex and confused, better resolution was obtained with a Philips camera of 11.46 cm. diameter; this camera was also used for lattice 52.

parameter measurements on terminal solid solutions.

The Nelson-Riley function was used to correct for systematic errors in the calculation of these lattice parameters.

The surfaces of a few alloys were examined using a camera in which a flat specimen could be rotated and oscillated through a predetermined angle; patterns obtained were directly comparable with those from a standard powder camera of 11.46cm. diameter.

Powder photographs for crystal structure determinations were obtained with a focussing camera of dia- eter equivalent to a standard 19cm. camera. Copper radiation and a quartz monochromator were used and the specimens were prepared by sprinkling the powders on to cellotape held rigidly to the same curvature as the film. A 0.1mm. scale was printed on to the film before it was developed, thus eliminating errors due to film shrinkage. The positions of the lines were measured with an illumin- ated fibre measuring device and the intensities were estimated visually.

The calculated intensities were obtained from the expression:- pp2 Ic O l+cos 2Gcos2 206 (l+cos2 20()sin2Ocos& 53. where c is the calculated intensity, e the Bragg angle, 4-the angle which the x-ray beam from the anti-cathode makes with the monochromater (for Cu K4radiation 20(= 26°40'), p is the multiplicity and F is the structure factor. The wavelength for Cu Kk1 radiation has been 0 taken to be 1.54050 A. 54

4. Results. Owing to the small amount of material available, no chemical analyses were carried out and, since weight losses on melting were generally low, nominal compositions have been accepted throughout. All compositions are given in atomic percentages. During heat treatment, some loss of thorium by oxidation to Zh02 occurred near the edges of the specimens. The oxide skin generally protected the specimens_ from the tantalum envelopes even when a considerable amount of liquid was present at the annealing temperature. Except where otherwise stated, melting points were estimated to -+ 12°C from the occurrence of incipient melting in annealed and quenched material after annealing up to 1500°G at 25°C temperature intervals. A sample from each two-phase region was annealed for seven days at 1000°C and, since only a few specimens were annealed at lower temperatures, the phase diagrams are accordingly only presented down to 1000°G. A list of the nominal compositions: of alloys prepared in each system is given in Appendix 2, together with a summary of the melting point data obtained by the incipient melting technique. Even when the melting points were above 1500°C, 55. many of the phase relationships could be deduced from the structures of as-cast and annealed samples and these results have been included in the phase diagrams. N,) attempt was made to determine the effect of the platinum metals on the f.c.c.(4) to b.c.c.(,) phase transformation in thorium and this has been assumed to be 1360±10°C(42) .

4.1. The thorium-ruthenium system. No difficulties were experienced during arc- melting alloys in this system and homogeneous buttons were obtained. Little oxide was present in the as- melted buttons; in alloys containing less than 505,6 ruthenium, it occurred as small dendrites while with more ruthenium it had a rosette-like appearance. In general, the x-ray patterns were clear which facilitated phase identification.

4.1.1. The phase diagram. These results are summarised in Fig.l. Four intermetallic compounds were detected; the crystal structures of three, Th7Ru3, ThRu and ThRu2, were determined completely, the other has been tentatively designated Th3Ru2 from the results of metallographic examination of as-cast alloys. 56.

There was a eutectic between thorium and Th Ru 7 3 at approximately 16% ruthenium and 1262±12°C. In polished sections, Th Ru tarnished rapidly to a golden- 7 3 brown colour and exhibited a characteristic striated appearance. Twelve months after mounting a sample of 30% ruthenium, the surface had corroded and become too uneven for metallographic examination. This compound crystallized from the melt at 1412±12°C and a eutectic occurred between Th Ru and Th Ru at approximately 37% 7 3 3 2 ruthenium; the eutectic temperature being 1388±12°C.

Th Ru was formed directly from the melt at 3 2 1425±25°C; it had a complex x-ray pattern which was not indexed and it did not respond to polarised light in polished microsections. There was a eutectic between Th Ru and ThRu; the eutectic composition was close to 3 2 Th3Ru2 and was estimated to be 41% ruthenium; the temperature was 1388±12°C. ThRu melted congruently at 1462±12°C; it occurred as a bright phase in freshly polished microsections but tarnished slowly on standing in air and responded strongly to polarised light.

ThRu melted above 1500°C and formed eutectics 2 with ThRu and ruthenium. The composition of the eutectic between ThRu and ThRu2 was 57% ruthenium and it melted at 1338±12°C. The second eutectic occurred at approxi- 57. mately 73% ruthenium and melted above 1500°C. ThRu2 did not respond to polarised light but had a rather greyer colour than ruthenium.

A specimen containing 1% ruthenium and a standard specimen of unalloyed, arc-melted, iodide thorium were solution treated together at 1000°C and then aged together at 600°C. The thorium specimen had a hardness of 48±1 VPN after solution treatment and this dropped gradually to 42.2±1 VPN during the 100 hours ageing; the hardness of the alloy remained 50±3 VPN throughout.

Metallographic examination showed that the alloy contained an appreciable amount of Th7Ru3 which occurred as a grain boundary eutectic network in the as-cast button and had balled up to form rounded particles after annealing for four days at 1000°C. There was no evidence of precipitation after ageing, either in the grain boundaries or within the grains. These results suggest that the solubility of ruthenium in thorium is very low at 1000°C and does not vary with temperature in the range 600-1000°C. Lattice parameters determined on powders taken from the thorium and alloy specimens after ageing 0 0 were a=5.0863-.0005A and 5.0869-1-.0005A respectively. This supports the low solubility of ruthenium in thorium since the difference in size between the thorium and 58. 0 ruthenium atoms would lead to a decrease of 0.0013A in the lattice parameter of thorium if 0.1 at% solubility of ruthenium occurred and Vegard's law were obeyed.

4.1.2.. The crystal structure of Th7Ru3. The crystal structure of Th7Ru3 was determined from a heavily exposed focussing camera film of an alloy containing nominally 30 at% ruthenium, a weight loss of 2% occurred during melting and subsequent metallographic examination suggested that the lost material was mainly thorium. Powder was obtained by crushing fragments of the as-melted button and it remained bright during the exposure of the film. The stronger lines of Th02 and patterns were all observed, the remaining theTh3Ru2 lines were indexed on a hexagonal unit cell a=9.969-.003A, c=6.302±.002A, a=0.63, calculated and observed values of singe are given in Table 2. The only systematic absences noted were hh 217 1 when 1 was odd.

The unit cell and systematic absences suggested that Th Ru might be isostructural with Th7Fe3 which has 7 3 4 a structure in the space group C6v P3mc. This structure would require two units of Th7Ru3 per unit cell, giving a calculated density of 11.82gm./cc. 59-

The positions of the atoms in this structure are: 2Thi in 2(b).3L. ,(, /3, 2./ zA.3,NiZe /3,1./ 21+z) \ z = 0,06 in 6(c) (x1 2x,z)(Tc;11 z)(xITC1 z)(75231;2+z)(2xt xy2i-z)(71x,2+z) 6Th11 x = .I26z = .25 6 in 6(e)(x, 2x, z)(2x, x, z)(x,x,2)(x1 2x,7-1-2-+z)(2x,x,i+z)(-1-I xti+z) Th111 x = .54z = .03 6Rh in 6(c)(x,2x, z)(2x,x, z)(xl x, z) (x, 2x,i+z)(2x,x,i+z) x = .815z = .31 These atomic parameters are the same as those given by Florio(23) for Th Fe • line intensities calculated on 7 3' the basis of this unit cell agree well with those estimated visually from the film as shown in Table 2.

The interatomic distances calculated on the basis of this structure are given in Table 3.

Four other compounds gave similar x-ray patterns, Th7Rh3, Th70s3, Th71r3 and Th7Pt3. Slight differences were observed between some of the line intensities in the patterns of Th7Ru3 and the last three compounds but these could all be explained by the differences in the scattering factors. The line intensities of these other compounds were not calculated but because of the similarity in the x-ray patterns the compounds were assumed to be isostructural. The lattice parameters and calculated densities are given in Table 4; observed and calculated values of sin20 are given in Table 5 and interatomic distances are given in Table 39 assuming that the atomic parameters are the same as those of Th7Ru3. 60.

4.1.3. The crystal structure of ThRu. An x-ray powder pattern of the alloy of 50 at% ruthenium was obtained with the focussing camera and was indexed on an orthorhombic unit cell a=3.878t.002 A, 0 a b=11.29t.01 A, c=4.0711.002 A. Calculated and observed values of sin261 ,Te cLown in Table 6. No hkl reflections were observed unless h+k was even and hOl lines were present only when 1 was even. The lattice parameters and systematic extinctions suggested that ThRu was isostructural with ThCo, CrB structure, space group 17 D2h Cmcm (No.63). The atomic positions are as follows:

4Th in 4(c) (01y1/4)(0,71 - 014-)(1/2,1h+Y1/4)(1/271h-Y7g)

4Ru in 4(c) ( " )( " )( )(

Line intensities calculated on the basis of this structure with -vTh . 0.140-0.004 and yRu = 0.410-0.004 gave good agreement with the observed intensities as shown in Table 6.

Three other compounds were found to have this structure, ThRh, ThIr and ThPt. At first sight the x-ray patterns of ThIr and ThPt did not appear similar to that of ThRu. This was found to be due to two contributary causes:- a) differences in the lattice parameters which led to different relative positions of 61.

several pairs of lines, b) the difference in the scattering factors of ruthenium and iridium or platinum was sufficient to alter the relative intensities of several weaker lines.

The lattice parameters, densities and atomic parameters are given in Table 7 calculated and observed sin26%and intensity values in Tables 8,9 and 10 and interatomic distances in Table 11.

4,1.4. The crystal structure of ThRu2. A focussing camera film of an alloy containing 66.6% ruthenium showed a simple pattern which was indexed 6 on the basis of a f.c.c. unit cell a=7.657-0.001 A. Calculated and observed values of sin2n are given in Table 12. Assuming that there are eight units of ThRu2 in each unit cell, the calculated density is 12.89 gm./cc. The observed intensities compare well with those for the MgCu2(C15) structure, space group Oh Fd3m, as shown in Table 12. The positions of the atoms in this structure are:-

8Th in 8(b) (//A)(7/8708)(74ACM08)(Y85/85/8)(AWOMAMAA) 62.

16Ru in 16(c) (000)(014/10(y0A)(0) (OW(Ogg)(AY4)(W1h) Ohgh)(0g )(Y40g)(M) (1020)(01/4)(4)(P/40)

Dwight, Downey and Conner(64) proposed that and ThIr also had the cubic MgCu (015) structure ThOs2 2 2 and this was confirmed by the results of examination of focussing camera films of alloys of 66.6% osmium and iridium. Lattice parameters of 7.715-0.001 and 7.6611 0 and ThIr respectively 0.001A were obtained for ThOs2 2 2n and the observed and calculated values of sin and line intensities are given in Table 13. Two weak lines not reported by Dwight were observed on each of the present films.

Nearest neighbour distances in ThRu2, ThOs2 and

ThIr2 are given in Table 14.

4.2. The thorium-rhodium system. During arc-melting, the metals alloyed easily and homogeneous buttons were obtained. Little oxide was observed after melting; it occurred as fine dendrites in alloys containing less than 66% rhodium and as small cuboids or asteroids with higher rhodium contents. In general, the quality of x-ray patterns obtained from 63. powders of alloys in this system were less clear than those from thorium-ruthenium alloys and required longer exposures.

4.2.1. The phase diagram. The phase diagram suggested for this s stem is shown in Fig.2. The crystal structures of Th7Rh3, ThRh, 3-ThRh2 and ThRh3 were determined completely; other compounds were observed at approximately 57, 62, 66 and 83% rhodium and these have been designated Th3Rh4, Th3Rh5,

,L-ThRh2 and ThRh5.

There was a eutectic between thorium and Th7Rh3 at 20% rhodium and 1237±12°C. Th Rh crystallized 7 3 directly from the melt at 1362±12°C and tarnished slowly in air; it formed a second eutectic with ThRh at 34% rhodium and 1312±12°C. The congruent melting point of ThRh was above 1500°C; this phase tarnished more slowly than Th Rh and had a strong pleochroic response to 7 3 polarized light. Th3Rh4 remained bright in air and was formed peritectically at 1487±12°C. It was soft and ductile and had a hardness of 155 VPN. The alloys containing 53 and 55% rhodium could be crushed only with difficulty and gave x-ray patterns of ThRh with a few extra lines. The alloys of 56, 57 and 58% rhodium could not be crushed and were filed. X-ray patterns obtained 64. on these alloys after the strain relieving anneal always gave a f.c.c. pattern with a = 5.085±.002 A which was identical with the lattice parameter of thorium metal. Poor quality x-ray powder patterns were obtained from the surfaces of microsections of these alloys in the as-cast state and after annealing at 1000°C and accounted for the extra lines on the films of 53 and 55% rhodium. No attempt was made to index this pattern.

Th Rh did not respond to polarised light and was 3 5 formed peritectically at 1450±12°C; at 64% rhodium it formed a eutectic with ThRh and this eutectic tempera- 2 ture was 1425±12°C. In the as-cast condition, a simple pattern was observed on x-ray films of alloys of 63, 64 and 65% rhodium, while specimens of 66 and 70% rhodium gave a complex pattern, (the films of 63, 64 and 70% rhodium also showed the presence of adjacent compounds). Results obtained on powders from annealed samples are shown below:-

Annealing temperature °C. % Rhodium As cast 600 1000 1200 1300 65 simple - - complex simple 66 complex complex complex - simple 65.

These results suggested that the simple structure ((3) was stable at high temperatures and transformed to the complex structure at low temperatures, the transformation temperature being above 1200°C. One possible explanation for the results obtained on as-cast alloys is that 5-ThRh2 can exist over a narrow range of composition and that the rate of the transformation from

0 is composition dependent and is lower at the thorium limit than at the rhodium limit of composition. Metallographic examination failed to assist with this problem since both the S.. and (3 phases were bright and tarnish resistant and both gave a good response to polarised light.

The ThRh peritectic horizontal extended little 2 on the thorium side of the compound since no spines of ThRh were observed in the as-cast ThRh2. ThRh 3 3 crystallized directly from the melt above 1500°C and was readily distinguished under the microscope by its characteristic orangy pink colour.

A greyish coloured compound, 'ThRh5 responded weakly to polarised light and was formed peritectically above 1500°C. In an as-cast alloy of 83% rhodium there were spines of ThRh3 surrounded by ThRh5 with small amounts of eutectic between the crystals; after 66.

annealing for one day at 1200°C, the spines of ThRh3 disappeared leaving an alloy which was almost single phased 'ThRh \ The 1ThRh ' peritectic horizontal did 5 ' 5 not extend to 85% rhodium. There was a eutectic between ThRh \ and rhodium at approximately 87% rhodium, the 5 eutectic temperature being 1450±12°C. Only poor quality x-ray films of 1 ThRh 'were obtained and no attempt was 5 made to determine the crystal structure of this phase.

The results of solubility and age hardening studies on a thorium-rich alloy were similar to those on the thorium ruthenium alloy, the hardness in this case remained 65±3 VPN and the lattice parameter after ageing was 5.0873± 0 .0005 A.

The crystal structures of Th7Rh3 and ThRh have been described in sections 4.1.2. and 4.1.3. respectively.

4.2.2. The crystal structure of 0-ThRh2. Powder from the as-cast alloy of 65% rhodium was examined by x-rays using the focussing camera. The pattern was indexed on a hexagonal lattice with

a=4.629-.002, c=5.849-.002 A and 7c = 1.264. Systematic extinction of the 0001 and hh2h1 reflexions were noted when 1 was odd and of hkil lines when h-k=3n and 1 was odd. The observed intensities showed excellent agree-

67.

ment with those for the Ni 2In structure (B82) space group D 6h P63/mmc as shown in Table 15 together with the observed and calculated values of singe. On this basis the density was 13.40gm./cc. In this lattice, the atomic positions are:-

2Th in 2(c) ( 3S 1/4)(3/3

2Rh1 in 2(a) (000)(0(W

2Rh11 in 2(d) (1/3 2/3 3/)(4 2/3 13 1/)4

and lead to the following interatomic distances:-

Th— 3Rh11 2.672 Rh1--2Rh1 2.924 Rh11 --3Th 2.672

2Ria11 2.924 6Th 3.046 2Th 2.924

6Rh1 3.046 6Riau. 3.046 6Rh13.046

6Th 3.960

4.2.3. The crystal structure of ThRh 3. Focussing camera films of powders from alloys containing 73, 75 and 77% rhodium were obtained in the as-cast condition and of the 73 and 77% specimens after annealing for 7 days at 100000. Each sample of the 77% alloy gave an x-ray pattern which could be indexed as simple cubic a = 4.148±0.003 A. Both films of the 73% 68. alloy showed weakly the stronger lines of old.i-ThRq'an.d another pattern which was similar to that obtained from the 77% rhodium alloy but with some of the lines occurring as doublets. The latter pattern was indexed on the basis of a tetragonal lattice a = 4.145±0.003, c = 4.111±0.003A, the same parameters being obtained both as-melted and after annealing at 1000°C. The as-melted 75% rhodium alloy also gave the tetragonal lattice with a = 4.142±0.0031 c = 4.111±0.003 A.

The patterns from material which had been annealed at 1000°C all showed the stronger lines of Th02 and the lines of ThRh3, both cubic and tetragonal, were less sharp than those on films of as-melted material. Metallographic- ally the 75% rhodium alloy appeared single phase while those of 77% rhodium were two-phase ThRh +ThRh ) 73 and 2 3 and (ThRh + ThRh ) respectively. A weight loss of 0.4% 3 5 occurred on melting the alloy containing 75% rhodium.

Intensities were calculated for the cubic lattice on the basis of the Cu3Au (L12) structure space group 1 0h Pm3m with the atoms in the following positions:-

1Th in 1(a) (000) 3Rh in. 3(c) (00600(6,004)0411h10)

69.

The agreement with the observed intensities is shown in Table 16 together with the observed and calculated values of sin28. Intensities for the tetragonal lattice were calculated on the assumption that it belonged to space group Doh -P4/mmm (No. 123) with the atoms in the following positions:-

1Th in 1(a) (000)

1Rh1 in 1(c) ( 11410) 1Th11 in 2(e) (0,1h1lh)(1h10,1h)

Good agreement with both the observed sin28 values and the line intensities was obtained as shown in Table 17.

The volumes of the unit cells of the ThRh phase in 3 the three alloys are given below and it will be seen that that of the cubic phase is larger than that of the tetragonal cell: Volume of % Rhodium 07 Lattice unit cell A' 73 70.6±0.2 Tetragonal 75 70.51:0.2 Tetragonal 77 71.4.0.2 Cubic

If all the lattice sites are occupied, there is no obvious reason why the structure should be tetragonal rather than cubic but the x-ray evidence suggests that 70. the compound can exist over a narrow range of composition and if the loss on melting the alloy of 75% rhodium were mainly rhodium, then the composition range might extend on the thorium side of the ideal composition. If up to one in 12 rhodium atoms in the (1410,14) and (0 1AM positions could be missing, this would allow a) a slight contraction in the "c" parameter, b) the compound to exist over the composition range 74.5 - 75% rhodium, c) the overall volume of the tetragonal cell to be up to 1.4% smaller than that of the cubic cell, i.e. it could account for the results obtained.

4.3. The thorium-palladium system. Thorium and palladium alloyed readily and no gross segregation was observed in arc-melted buttons. In alloys containing more than 80% palladium, there was a tendency for small droplets to break off after the initial mixing had taken place, but as far as could be ascertained by metallographic and hardness tests, these droplets had the same composition as the parent buttons. They might have been caused by the release of residual gases absorbed by the palladium. 71.

4.3.1. The phase diagram. The results obtained on this system have been summarized in Fig.3. Seven intermetallic compounds occur, the structures of Th2Pd ThPd and ThPd , Th3Pd5' 3 4 have been fully determined; the other compounds occurred at 50, 57 and 83% palladium and have been called ThPd, Th3Pd4 and ThPdX.

4.3.1.1. Alloys of 0-75% palladium. Pd A eutectic was formed between thorium and Th2 at approximately 23% palladium and 1112±12°C. and Th2Pd melted congruently at 11621:12°C. This compound always appeared dark on microsections and the fracture surfaces of buttons rich in Th2Pd were also dark. It oxidised in air and small lumps in a sealed glass bottle changed in a few weeks to a black powder which gave a diffuse x-ray pattern on which the lines of Th02 could just be distinguished. Th2Pd formed a second, coarse eutectic with 'ThPd at 36% palladium and 1137±12°C. The eutectic structure was observed in alloys on the thorium side of the eutectic composition but not in alloys con- / taining primary ThPd unless the specimens were cooled very slowly. This was possibly because the /Tha` of the eutectic nucleated on the primary dendrites of iThPe. ThPd‘ melted at 1412±12°C and tarnished in air less 72.

Pd but powder sealed in an x-ray rapidly than Th2 capillary gave only a diffuse pattern of Th02 one month after preparation. Only a poor quality focussing camera film of / ThPd\ was obtained and no attempt was made to determine the structure of this phase.

The compound i Th3Pd14.\ was formed peritectically at 1325±12°C. The temperature was deduced from the fact that two specimens in the (/ThPd` + /Th3Pd4 phase field were annealed at this temperature and one remained unmelted while the other showed partial liquidation. The focussing camera film ofThPd % was underexposed and of poor quality and no attempt was made to measure the film but the pattern appeared to be similar to that of ITh3Pt4

Th Pd I formed a eutectic with Th Pd at 1212±12°C 3 4 3 5 and approximately 62% palladium; the eutectic structure was sometimes difficult to detect in as-cast specimens since the small amount of Th3Pd4 present was distri- buted as very fine particles. Th3Pd5 was formed peritectically at 1387±12°C and the peritectic horizontal did not extend to 62% palladium.

The compound ThPd3 melted congruently at a temperature above 1500°C and no grain growth or rounding 73. of the crystal edges occurred on annealing at this temperature. In alloys containing an excess of thorium, ThPd3 occurred as bright, well developed crystals which often had a hexagonal tendency in microsections and showed a good response to polarized light; the crystalline form was less well marked in alloys containing an excess of palladium. A similar effect was noted by Catterall(71) for the compound UPd3 in the uranium-palladium system.

The resistance of the compounds Th2Pd, 'ThPd,' Th Pd Th Pd and ThPd to atmospheric oxidation 3 4' 3 5 3 increased with the palladium content of the phases which were readily distinguished in polished microsections.

Although the structure of the compound which came into eauilibrium with thorium in this system was different from that of the thorium-rich compound in the thorium-ruthenium and thorium-rhodium systems, similar results were obtained for solid solubility and age- hardening studies on all three systems. The hardness of the 1% palladium alloy remained 54±3 VPN during ageing and the lattice parameter at the end of this 0 treatment was 5.0873±0.0005 A. 74.

4.3.1.2. Alloys of 710 palladium. Difficulties were experienced in positioning the solidus boundary in this composition range since a fine eutectiferous grain-boundary network was sometimes observed after annealing at temperatures above 1000°C. The effect was thought to be due to contamination since it generally did not spread right through the specimen and the amount of grain-boundary network did not appear to depend on composition or annealing temperature in any rational manner.

Annealing and quenching studies showed that there was a sharp minimum in the solidus at approximately 12% thorium and 1125±12°C. Coring was observed in as-cast alloys containing 5-10% and 1656 thorium but not in those of 12-14% thorium; these results are consistent with the solidus minimum at about 12% thorium. No discon- tinuities in the solidus were detected between 12 and 21% thorium at which composition the solidus temperature was above 1350°C.

It was thought at first from metallographic results that thorium was soluble in palladium up to about 20% and that this solid solution came into equilibrium with ThPd3. However, x-ray results on powders annealed at 1000°C showed that the limit of solubility was lower. 75-

The lattice parameter of f.c.c. palladium increased approximately linearly up to 15% thorium and the values are given below:- 0 Nominal % Thorium 'a' in A 0 3.890±0.001 5 3.945±0.001 11 4.036±0.001 15 4.053±0.001

The pattern of the 17% thorium alloy appeared to be related to the f.c.c. pattern of the 15% thorium alloy since the stronger lines occurred at similar positions; however, some of these stronger lines had split and other lines were also present indicating that the lattice was more complex. It was not possible to index the film on a simple variant of the cubic lattice. This pattern was strikingly similar to that obtained at AERE by Polls(72) on a uranium-palladium alloy of similar composition. The complex pattern was also obtained after annealing and quenching the alloy of 17% thorium from 1150°, 1100° and 800°C. An alloy of 19% thorium also gave the complex x-ray pattern after annealing at 1000°C but the positions of the lines indicated a slight increase in unit cell size with thorium composition. The 20% thorium alloy gave a simple cubic pattern after annealing at 1000° and 1200°C. For convenience, the 76. phase giving the complex pattern has been called ThPdx and that with the simple cubic pattern ThPd4. Phases present in alloys of 15-23% thorium are summarized in Table 19.

In metallographic specimens, ThPd3 was recognised by its very bright colour, and ThPd41 ThPdx and palladium could not be distinguished from each other. They were all slightly less bright than ThPd3 and did not respond to polarised light; no satisfactory etch was obtained, concentrated HNO brought up the grain boundaries but did 3 not differentiate between the phases. A considerable amount of further work is needed in order to establish full details of the phase diagram in this composition range.

The solution of thorium in palladium was accom- panied by an increase in hardness, little difference was found between specimens in the as-melted condition and those examined after homogenizing for four days at 1000°C even when coring was removed by the anneal. The results are given in Table 19 and show that the maximum hardness of 547 VPN was observed at 17% thorium, the composition at which the x-ray pattern was no longer simple f.c.c. Alloys containing more than 17% thorium tended to be brittle and to crack when the load was applied. 77. 4.3.2. The crystal structure of Th2Pd. (68) Ferro has stated that this phase has the CuAl2 18 structure, space group D 4h I/mcm and, in confirmation of this, a focussing camera film of powder from an as- melted alloy of 33.3% palladium gave an x-ray pattern which could be indexed on a tetragonal lattice with a = 7.308±0.002, c = 5.960-1-0.003 A and i = 0.815, observed and calculated values of singe are given in Table XX.

Intensities were calculated on the basis of the CuAl2 structure with the atoms in the following positions:-

8Th in 8(h) (xl1/44-x10)(Fil x,0)(1h+x,R10)(Y2-x,x,0) (4+x,x1 11h)(h-x1R,lh)(illhi-x114)(x11/2-xl1/4) x=0.158 4Pd in 4(a) (oog.)(oog)(veg)(01)

The calculated density was then 11.90 gm./cc. and the interatomic distances are:- 0 0 Th - 4Pd 3.10 A Pd - 2Pd 2.98 A 1Th 3.26 8Th 3.13 2Th 3.54 4Th 3.77 4Th 3.89 78.

4.3.3. The crystal structure of Th3Pd5. The focussing camera film of the alloy of 63% palladium was indexed as hexagonal with a=7.149±0.003, C C c=3.899±0.002 A. g = 0.55 and observed and calculated values of sin26 are given in Table XXI. No systematic extinctions were observed. The volume of the unit cell suggested that the compound was Th Pd with one unit of 3 5 Th Pd per unit cell; this gave a calculated density of 3 5 11.84 gm./cc.

A structure with:-

in 5,2/j,0; 2 Pd1 3 Pd11 in xpd10,0; O,XPd,O; 5Epa,Rpd70 with xpd =0.75 3 Th in xTh,0,14; 0,xTh114; 7Th,2Th,lh with xTh =0.33 gave a rough fit to the observed intensities. The atomic parameters were refined using the A.E.R.E. Ferranti 'Mercury' computer, the refined parameters xpd =0.780:10.002 and xTh = 0.350±0.002 gave good agreement between observed and calculated intensities as shown in Table XXI. The highest symmetry space group in which these atomic positions are possible is D3h P62m (N° 189). Interatomic distances for Th Pd were calculated to be as follows:- 3 5

79. 0 0 2.93A Pd1 - 3Pd11 2.87A Pd11 - 2Pd11 2.72A Th 4Pd11 6Th 3.03 2Pd1 2,87 4Pd1 3.03 4Th 2.93 2Pd11 3.64 2Th 3.64 2Th 3.93 4Th 4.03

The focussing camera x-ray film of the alloy of 63% platinum was similar to that of Th Pd and was indexed on 3 5 the basis of a hexagonal lattice with a=7.162±0.003 and , c=3.908±0.002 A = = 0.55. No attempt was made to refine a the atomic parameters of this compound and consequently it is not possible to give calculated intensities or inter- 2 atomic distances. Observed and calculated values of sin 0 are given in Table XXII.

4.3.4. The crystal structure of ThPd3. The structure of this compound was determined from a focussing camera film of powder from the as-melted alloy containing 75% palladium. The pattern was indexed as 0 hexagonal with a = 5.858±0.003, c = 9.814±0.003 A and

-a =1.67, observed and calculated values of sin26 are given in Table XXIII. The only systematically absent reflexions were hhl when 1 is odd and the observed intensities -show good agreement with the TiNi structure, space group 3 D6h P63 /mmc, as shown in Table XXIII. This structure 80.

per unit cell giving a calcu- requires four units of ThPd3 lated density of 12.60 gm./cc. The atomic positions are:-

(000)(001h) 2T h1 in 2(a) 2Th11 in 2(c) (;'3 3' 4/)( 2, 3/ I/)3 6Pd1 in 6(g) (1200)(01,20)(//0)(0WM1/4)() 71<)(27,37,,1<)(x,X,)(7,25E,Y) 6Pd11 in 6(h) (x,2x (2x,x,34)(7,x34) and the intensity agreement was obtained when xpa had the ideal value of The distances between nearest neighbours were then:- 0 0 2.92 A Th - 2.92 A Th1 - 6Pd1 11 6Pd11 2.98 6Pd11 2.98 6Pd1 6Th 4.18 6Th11 4.18 1 0 0 Pa - 2.92 A Pd1 - 4Pd1 2.92 A ll 4Pd11 2Th 2.92 2Th1 2.92 11 4Pd1 2.98 4Pd11 2.98 2Th 2.98 2Th11 2.98 1

ThPd3 was found to exist over a narrow range of composition and lattice parameters, axial ratios, unit cell volumes and closest Th-Pd neighbours are given below. The stronger lines of Th3Pd5 were also observed on the film of the 70% palladium alloy and the ThPd4 pattern occurred on that of 78% palladium. 81.

0 0 Volume of Closest unit cell Th-Pd %Pd "a" A "c" A a 0 7 A' neighbours 70 5.8601.003 9.8061.004 1.67 291.6 2.930 75 5.8581.003 9.8141.004 1.67 291.6 2.929 77 5.851±.003 9.657-1-.004 1.65 286.3 2.925 78 5.846±.003 9.6431.004- 1.65 285.4 2.923

4.3.5. The crystal structure of ThPd4. The film from the alloy of 80% palladium was indexed 0 as simple cubic with a = 4.110-1-0.002A and observed and calculated values of sin20 are given in Table XXIV. Intensities were calculated on the basis of a Cu3Au-type lattice with one in five of the thorium atom sites occupied by palladium atoms and good agreement with the observed intensities was obtained as shown in Table XXIV. The compound existed over a narrow range of composition and the variation of lattice parameter, unit cell volume and closest Th-Pd neighbour distance with composition are shown below. 0 %Pd. "a" A Volume ofo Closest Th-Pd unit cell A3 neighbours 78 4.1261.001 70.24 2.918 79 4.119-1.001 69.88 2.913 80 4.1101.001 69.43 2.905 82.

Visual estimation of the line intensities was sufficiently sensitive to distinguish between the possible defect structures for this lattice, e.g. thorium sites vacant or occupied by palladium atoms, since both gave only slight alteration in the relative intensities between the f.c.ce lines and the superlattice lines. Random substitution for thorium atoms is suggested as this allows for the phase to exist over a range of composition with an increase in the volume of the unit cell with increasing thorium composition as observed.

In Cu3Au-type structures when, as in ThPd4, the size difference between the atoms is appreciable, the co-ordination around the larger atom is increased to 18 (See Section 2.2.7) by the addition of six neighbours in corresponding positions at a distance of "a". In the non- ideal structure proposed for ThPd4 the average co-ordination for thorium would be 4.8Th +(12 + 1.2)Pd = 18 and for palladium 3.2Th +(8 + 0.8)Pd = 12 at 80% palladium.

4.4. The thorium-osmium system. Difficulties were experienced in obtaining homogeneous ingots in the composition range 35-50% osmium and in spite of repeated meltings, lumps of unmelted osmium were observed frequently on subsequent metallographic examination. Homogeneous melts were obtained in alloys of 83.

0-35 and 50-100% osmium. Good quality x-ray patterns were generally obtained from thorium-osmium alloys.

4.4.1. The phase diagram. The results for this system are summarized in Figure IV. Three intermetallic compounds were observed and the crystal structures of Th 0s and ThOs have been 7 3 2 fully determined while the third compound which occurred between 36 and 50% osmium has been designated ThOsx.

Eutectics occurred between thorium and Th 0s and 7 3 between Th 0s and ThOs at 13% and 36% osmium and the 7 3 x eutectic temperatures were 1287±12 and 1487±12°C respectively. Th70s3 and ThOs2 were both formed directly from the melt at temperatures above 1500°C while ThOsx was formed peritectically and again the melting point was above 1500°C. There was a eutectic between ThOs2 and osmium at approximately 85% osmium and this temperature was also above 1500°C. The x-ray pattern of ThOsx was complex and no attempt was made to determine the crystal structure of this phase.

All three intermetallic compounds in this system remained bright on polished microsections; ThOsx had a slightly yellowish tinge in ordinary reflected light and

ThOs2 was distinguished by the fact that it did not 84. respond to polarised light in the as-polished condition while Th 0s and ThOs gave good response. 7 3 x

Age hardening studies were made on a 1% thorium- osmium alloy and showed that at 600°C the alloy did not age harden and the hardness remained 58.2t2 VPN. Metallographic examination suggested that the solid solubility at 1000°C and at 600°C was considerably less than 1% and, in confirmation, the lattice parameter after 0 ageing was 5.0873±0.0005 A.

The crystal structures of Th70s3 and ThOs2 have been described in sections 4.1.2 and 4.1.4 respectively.

4.5. The thorium-iridium system. No difficulties were encountered during arc-melting alloys in this system and homogeneous buttons were prepared. On x-ray examination, very good, well resolved patterns were obtained from both as-melted and annealed specimens. Little oxide was observed in as-melted alloys, it occurred as small dendrites in alloys containing less than 50% iridium and as small asteroids or cuboids with more iridium.

4.5.1. The phase diagrRm. Figure V summarizes the results obtained on this system. The crystal structures of ThIr and ThIr Th7Ir3' 2 85. have been fully determined and have been described in Sections 4.1.2, 4.1.3 and 4.1.4 respectively. Dwight has (43) proposed a structure for ThIr . Evidence for two 5 additional compounds was obtained, one occurred between Th71r and ThIr and the other at 75% iridium and these 3 have been called ThIrx and 'ThIr '.

71r at There was a eutectic between thorium and Th 3 approximately 15% iridium and 1337±12°C. Th71r3 and ThIr crystallized directly from the melt at temperatures which were greater than 1500°C. and both responded to polarised light in polished microsections, the response by ThIr being stronger than that by Th71r3. A eutectic structure was observed in an as-cast alloy of 35% iridimn; in the as-cast alloys of 37 and 40% iridium there was evidence of the peritectic formation of another compound around some of the dendrites of ThIr but after annealing at temperatures in the range 1000-1450°C only ThIr and Th7Ir3 were observed metallographically. On the x-ray pattern of the as-cast 40% iridium alloy, the stronger lines of Th71r3 and ThIr were all observed together with several extra lines, some of which were strong. X-ray examination after annealing between 1000° and 1450°C confirmed that only Th71r3 and ThIr were present.

The x-ray pattern of the as-cast alloy of 33% 86.

iridium showed the presence of Th71r and the phase present 3 in the as-cast alloy of 40% iridium. After annealing and quenching the alloy of 37% iridium from 1475°C and the alloy of 40% iridium from 1475° and 1500°C, areas of quenched liquid containing small dendrites of a phase which was less bright than Th71r or ThIr and a fine eutectic were observed 3 and in the alloy of 40% iridium larger crystals of the darker phase which had not been molten at the annealing temperature were also observed. The specimen of 40% iridium which had been quenched from 1500°C gave an x-ray pattern which contained the extra lines present in the as- cast specimen. It is suggested therefore that a compound

ThIrx was formed slowly, by a peritectic reaction between ThIr and liquid and decomposed eutectoidally to ThIr and Th71r at a temperature between 1450° and 1475°C. Since 3 the as-cast alloy of 33% iridium gave the patterns of Th71r and ThIr it is suggested that the eutectic at 35% 3 x iridium melted at 1462±12°C and was between Th71r and 3 ThIrx and that the decomposition temperature of ThIrx was just below the eutectic temperature; (on Figure V this difference has been exaggerated for clarity). The ThIrx peritectic horizontal extended to approximately 36% iridium and the composition of the compound was close to 40% iridium. 87.

All alloys of more than 50% iridium melted above 1500°C. There were eutectics between ThIr and ThIr and 2 between ThIr and iridium at approximately 57 and 90% 5 iridium respectively.

No phase relationships are shown in Figure V for the composition range 66-83% iridium since all the alloys were bright and tarnish resistant and no suitable etchant was found. :detallographic evidence for the existence of the phase 'ThIr3' is based on the fact that ThIr2 did not respond to polarised light while alloys of 75% iridium responded completely, the responsive phase increasing in amount between 68 and 75% iridium and all alloys of 75-83% iridium responded completely. No eutectic structures were detected in this composition range under polarised light. and between Since eutectics occurred between ThIr and ThIr2 ThIr5 and iridium, one of the phases ThIr2, ThIr3 or ThIr5 must crystallise directly from the melt while the other two wore probably formed peritectically.

X-rayevidencefortheexistenceof i ThIr-‘was complicated by the fact that there was a great deal of similarity in the positions of lines of the three compounds \ 2 ThIr2, ThIr and ThIr but in Table XXV the sin 0 values 3 5 from films of 66,72,75,80 and 83% iridium are compared and it will be seen that the films of 72 and 80% iridium 88. , and contain the stronger lines ofThIr( 2 and ThIr3 (iThIr and ThIr5) respectively. Owing to the simi- 3 larities in the positions of these stronger lines it is possible that the structure of ThIr is closely related 3 to those of ThIr and ThIr although no attempt was made 2 5 to determine its structure.

A 1% iridium alloy was aged at 600°C but no increase in hardness was observed and the hardness remained 58.2± 2 VPN. The solid solubility of iridium in thorium at 1000° and 600°C was appreciably less than 1% from the results of metallographic examination and the lattice parameter after ageing was 5.0873±.0005 A.

4.5.2. The crystal structure of ThIr5. A focussing camera film was obtained of powder from the as-cast alloy of 835fi iridium. The pattern was more complex than expected on the basis of the hexagonal structure proposed by Dwight (CaZn5)(43). Many of the lines relating to his structure were doubled and several extra lines could not be accounted for. If average sin20 values were taken for the lines close to the positions expected from Dwight's structure then they could be indexed reason- ably well on a hexagonal lattice with a=5.335±0.003, c c=4.286±0.003 A, —L=0.80a these parameters are close to those

89.

proposed by Dwight. On this basis there are no systematic absences, which is consistent with space group D6h1 P6/mmc to which CaZn5 belongs.

Since some lines were not doubled and the relative intensities of the first and second lines of a pair were not consistent, e.g. lines 200 and 111, it seems that the deviation from Dwight's structure is genuine and was not due to inhomogeneity in the specimen. The extra lines could not be accounted for on a simple variant of this structure such as doubling one of the lattice parameters. An identical x-ray pattern was obtained from material which had been annealed for one day at 1200°C.

Observed and calculated sin2 (3 values and intensities calculated for the CaZn structure are given in Table 5 XXVI. Interatomic distances calculated on the basis of the CaZn structure are:- 5 0 2.64 Th-12Ir1 3.07 A Ir1-6Ir11 2.64 Ir11-4Ir1 41r 2.65 61r11 3.415 31r1 3.07 11 2Th 4.29 3Th 3.07 4Th 3.41

4.6 The thorium-platinum system. During the arc-melting of thorium-platinum alloys, it was observed that the temperature of the metals rose appreciably as soon as they ran together suggesting that 90. a strongly exothermic reaction occurred. This effect was noted over the whole range of composition and resulted in homogeneous buttons. Very good quality x-ray patterns were generally obtained from these alloys.

4.6.1. The phase diagram. The phase diagram proposed for this system is shown in Figure VI. Eight interinetallic compounds were detected and the crystal structures of three, Th7Pt3, ThPt and Th Pt were fully determined. Other compounds have been 3 5 / / / called Th Pt. Thlt ThPt ThPt and ThPt as a 3 4 , 2 , 3 ' 4 5 result of metallographic and x-ray examination of as-melted and annealed specimens.

Thorium and Th Pt formed a eutectic at 17% platinum 7 3 and this melted at 1237±12°C; Th Pt was formed by a peri- 7 3 tectic reaction at 1362±12°C and the peritectic horizontal did not extend to 25% platinum. Both of these temperatures were confirmed by differential thermal analysis when very strong differential emf effects were observed during melting and freezing.

Between 50 and 83% platinum, the solidus temperatures were above 1500°C and were not determined. ThPt, Th Pt5 and ThPt crystallised directly from the melt 3 3 while Th Pt Thlt ThPt and ThPt formed 3 4 , 2 , 4 5 91. peritectically. There was a eutectic at 53% platinum 1 between ThPt and / Th3Pt4 and one at 69% platinum between i \ / \ / i ThPt and ThPt . The eutectic between ThPt and 2 3 5 platinum occurred at 90% platinum and melted at 1337-1:12°C.

Th Pt and ThPt both tarnished to a brown colour on 7 3 standing in air but could be readily distinguished in polished microsections since Th Pt tarnished more rapidly 7 3 and ThPt had a very strong pleochroic response to polarised light. Difficulties were experienced in distinguishing metallographically between phases containing more than 50% platinum; they were all bright and tarnish resistant and all showed some response to polarised light. The latter effect enabled the relationship between the phases to be established while the individual phases were readily differentiated on x-ray powder patterns.

1ThPt was less brittle than the other compounds 5 and had a hardness of 593-1:8 VPN.

The x-ray patterns of the alloys of 57, 66, 75, 80 and 83% platinum were not particularly complex but no attempt was made to determine the crystal structures of these phases. It was noted that the patterns of the alloys of 57% palladium and 57% platinum were very similar.

An alloy of 1% platinum was aged at 600°C and the 92. hardness at room temperature remained 53.2±3 VPN. throughout. Metallographic examination after solution treatment and after ageing suggested that the solid solubility of platinum in thorium was appreciably less than 1% at both temperatures and the lattice parameter after ageing was 0 5.0942±.0005 A.

The crystal structures of Th7Pt31 ThPt and Th3Pt5 have been described in sections 4.1.2, 4.1.3 and 4.3.3. respectively. 93.

5. Discussion 5.1. The phase diagrams. 5.1.1. Accuracy of the results. On economic grounds, only one gm. alloy samples were used and it was not possible for any chemical analyses to be performed on this amount of material. As described in section 3.1. weight losses on arc-melting were generally low and it is considered that the compositions of the eutectics and intermetallic compounds are accurate to -1% except for ThOsx and ThIrx for which the especial difficulties have already been described. Since the main emphasis in this work has been to obtain the maximum amount of information on the general form of the phase diagrams, 25°C temperature intervals were used in the annealing programme giving the fixed temperatures to ±12°C. Every fixed temperature was confirmed by annealing at least two samples at temperatures such that one sample showed liquation and one did not. In a few cases the repeat anneals gave one sample which showed liquation at a particular temperature and one which did not at the same temperature. In such cases the fixed temperature has been taken as the annealing temperature and it has been assumed that the variation in results is due to errors in thermo- metry. 94.

5.1.2. Points of similarity between the systems. Since the present experimental techniques gave an upper annealing temperature of 1500°C, it proved possible to determine more fixed temperatures in the phase diagrams of thorium with ruthenium, rhodium and palladium than in the systems with osmium, iridium and platinum but some general points have emerged:-

a) In each series, the first element (ruthenium and osmium) forms fewer compounds and therefore has a simpler phase diagram than the second or third elements.

b) Thorium forms a eutectic with the thorium-rich intermetallic compound in each system.

c) On a 'Raynor' plot of these eutectic temperatures vs the melting points of the compounds, the points lie close to a straight line which includes similar points from the systems of thorium with copper, nickel, indium, aluminium and selenium.

d) In the systems of thorium with ruthenium, rhodium and palladium, a high melting point compound occurs in the range 66-75% transition element and there is a subsidiary maximum in the liquidus at the equiato.mic composition.

e) The system Th-Pd has the fewest points of similarity with the other five systems. 95.

f) Except for Th2Pd, the compound richest in thorium in each system belongs to an isostructural family Th7X3.

g) Except for ThPd, the compounds ThX belong to an isostructural family. (There is no compound of this composition in the Th-Os system.)

h) Except for Th-Pd there is a eutectic between the transition element and an intermetallic compound in each system.

i) Except for the solubility of thorium in palladium, terminal solubility is restricted and no significant age- hardening of thorium-rich alloys was observed.

5.1.3. The thorium-rich eutectic temperatures. The occurrence of compounds has a large influence on the shape of an equilibrium diagram and on the composition and temperature of any neighbouring eutectics. If the compound is very stable and thus high melting, a considerable amount of the solvent must be added in order to modify the thermodynamic activity of the solute until it equals that of the solvent-rich liquid. Raynor(73) has pointed out that, for magnesium alloys, the higher the melting point of the first magnesium-rich compound, the higher the eutectic temperature will be and the lower the eutectic composition. 96.

Waber and Gschneidner(12'13) have shown that similar effects occur in alloys of plutonium, lanthanum, cerium and praseodymium. This 'Raynor' effect is examined for thorium-rich eutectics in Figures VII and VIII and it will be seen that, in general, the eutectic temperatures increase and the eutectic compositions decrease as the melting points of the compounds increase. The data for beryllium appears to be an exception to this general rule but this is the only system in which the compound contains less than 50% thorium.

As a result of an examination of the factors affecting the 04/(3 phase boundaries of zirconium and titanium alloys, Betterton and Frye(74) noted that a plot of a common convenient temperature such as a eutectoid temperature vs electron concentration indicates that the temperature decreases as the electron concentration decreases. Waber and Gschneidner extended this idea to cerium-rich eutectic alloys(13) and found that a similar trend occurred. In Figure IX a similar plot is shown for the thorium-rich eutectics. A valence of 0 was assumed for the group 8 metals, 1 for copper, 2 for beryllium, 3 for aluminium and indium, 4 for thorium and 7 for selenium. The mean electron concentration was obtained from the mean of the electron concentrations at the liquidus and at the 97. solidus. It will be observed that an approximately linear relationship is obtained, that the point for beryllium lies close to this line and that the points for copper, indium, aluminium and selenium lie below the line.

5.1.4. Melting...points of the compounds. High melting points are generally assumed to be an indication of high heats of formation and hence of high stability of intermetallic compounds. It follows therefore, that in the systems of thorium with ruthenium, rhodium and palladium the maximum stability occurs at compositions corresponding to ThRu2, ThRh3 and ThPd3. In an examination of available information on the alloy systems of aluminium with titanium, zirconium, thorium, uranium, (1) plutonium, lanthanum and cerium, Murray pointed out that there was a high stability of the intermetallic compounds which occurred at 66-75% aluminium. This high stability was not associated in any obvious way with volume concentration on formation nor with a particular crystal structure.

(2) Laves has pointed out that in certain (AB2) Laves phases there is correlation between the observed contraction in the A component and the melting point of the compound and that a high contraction tends to be associated with a high melting point. Since the melting points of 98. many of the present compounds are still unknown, it is not yet possible to assess how far this correlation can be generalised.

5.1.5. Age-hardening- behaviour and the lattice parameter of thorium. The equipment and techniques used in the age- hardening tests were the same as those used in earlier studies(30,76) of the age-hardening of thorium-rich Th-Al, Th-In and Th-U alloys, so that a direct comparison with those results is possible. Th2Pd is isostructural with the thorium-rich compounds in the systems with aluminium and indium and thorium-rich alloys from both of these systems showed a significant increase in hardness on ageing at 600°C although the solubility of aluminium in thorium was only 0.4% at 1000°C (the homogenising temperature).(75) The fact that Th-Pd alloys did not harden on ageing supports the metallographic evidence that the solubility of palladium in thorium is low and does not change between 1000° and 600°C. The intermetallic compounds Th2X in these alloys are isostructural with CuAl2 and-thorium, like aluminium, is f.c.c. It is therefore possible that in the thorium systems where a change in solid solubility with temperature occurs, the precipitation might be multi-stage, as is the case with aluminium-rich alloys in the Al-Cu 99.

system. If this is so, then one stage might be the formation of a coherent precipitate with consequent increase in hardness.

In the five alloy systems where a thorium-rich com- 7X occurs, the solid solubility is pound of the type Th 3 again very low at both the homogenising and the ageing temperatures. At the boundaries between the thorium matrix 7X particles a considerable degree of and the complex Th 3 misfit must occur and as with Th-U alloys a coherent precipitate cannot form. A lack of age-hardening in the five Th7X systems is therefore in accordance with results 3 on other thorium systems.

The slight increase in hardness of the 1% palladium and 1% platinum alloys over the hardness of unalloyed thorium is in agreement with the results of Goldhoff, Ogden and Jaffee(77) who examined the room temperature hardness of a series of binary thorium alloys which. had been annealed for two hours at 850°C. Their value for unalloyed thorium was 80 171T suggesting that their material was less pure than the present material and they obtained hardness increases to 85 and 83 VPN on the addition of 0.5% palladium and platinum respectively.

The lattice parameter obtained for unalloyed 100. 0 iodide thorium after ageing at 600°C was 5.0863-0.0005 A which is slightly higher than the value given by Evans and Raynor(78) (5.0842±0.0002 A) for arc-melted samples of the same starting material. The latter value was obtained after extreme precautions had been taken to avoid contamination during the preparation of the thorium powder. In view of the results of Evans and Raynor it is probable that the difference in the two values is due to slight contamination during the preparation of the present material. This suggestion is supported by the fact that the lattice parameter of thorium powder which had been strain-relieved at 1000°C was appreciably higher. The lattice parameters of the aged 1% alloys differed little from that of the unalloyed thorium and supported the metallographic evidence that the solubility of these elements in thorium was low. Therefore it was not considered worth while to repeat the x-ray work using modified specimen preparation techniques.

5.2. The intermetallic colpounds. 5.2.1. Compounds of the type Th . 7 X3 The lattice and atomic parameters given by Ferro and Rambaldi(66) for Th Rh and by .qatthias et al.(67) for 7 3 Th7Os and Th_Ir., are in excellent agreement with the - 3 values obtained in the present work. The radius ratios of thorium to the platinum group metals lie in the range 101.

1.30-1.34 and are thus within the range for other compounds having this structure (1.30-1.47). Extremes of axial ratio reported for this structure are 0.624 for Th7Fe3 and 0.651 for Re7B3 and it will be seen from Table IV that the axial ratios of the five new compounds lie within this range. It can also be seen from Tables III and IV that, a) the corresponding interatomic distances and, b) the volumes of the unit cells, increase with the atomic diameter of the X element.

In their discussion of the crystal structure of (24) Ce7Si Roof et al. pointed out that in the paper by 3' Florio et al.(23), Table I, which lists the interatomic distances in Th7Fe3, is in error with respect to both the distance given and the identification of neighbours. The present author agrees with Roof et al, thus explaining the apparent discrepancy between interatomic distances etc. in Th7Fe and in the five new compounds. 3

5.2.2. Compounds of the type Th2X. (68). Pd was published by Ferro The structure of Th2 while the present results agree that the compound is isostructural with CuAl2, the two sets of lattice parameters differ significantly and lead to a considerable difference in the relative separations of several pairs of lines. 102.

The present author did not obtain any evidence for a range of composition of this phase but her alloys were examined only in the as-cast state whereas Ferro's samples were all annealed at several temperatures with the final anneal at 500°C.

The radius ratio of thorium to palladium (1.31) is within the range given by Laves for this structure to be (27) stable . Seven compounds of thorium are now known to crystallise with this structure, namely Th2Ag, Th2Al, Th2Au, Th2Cu, Th2In, Th2Pd, Th2Zn. In Figure X the electro- c(27) negativity of the X element, as given by Lave.„ 9 has been plotted against the atomic diameters for C.N.12, as given by Pauling(79), and the closest Th-X distances observed in these compounds have been written on the plot. It will be observed that:-

a) for a given diameter of X the Th-X distance decreases as the electronegativity increases,

b) for a given electronegativity of X, the Th-X distance decreases with the diameter of X.

(28) This is an extension of the observations of Murray which were based on an analysis of the interatomic distances in Th2A1, Th2Ag, Th2Au and Th2Cu. It is interesting to speculate on whether any other isostructural compounds of 103. thorium would occur; from a study of the periodic table, Th Ga and Th Tl appear possible and on the plot of 2 2 Figure X the Th-Ga distance would be expected to be greater than 3.22 while the Th-Tl distance should lie 0 between 3.20 and 3.31 A.

5.2.3. ThX compounds with the CrB (Bf) structure. Six compounds of thorium are now known to have this structure; they are ThAl, ThCo, ThRu, ThRh, ThIr and ThPt. (34) Aronsson has noted that the occurrence of this structure depends to a rather large extent on a favourable radius ratio r„ = 1.43 and in the present series of com- rX pounds this ratio varies between 1.45 for ThCo and 1.26 for ThAl in general agreement with Aronsson's observations. The volumes of the unit cells increase with the diameter of the X atoms as shown in Table VII. In this structure the X atoms form zig-zag chains through the lattice and the X-X distances within these chains are also shown in Table XI together with the other interatomic distances.

Apart from ThAl, the shortest X-X distances show an increase of approximately 10% and the shortest Th-X distances a decrease of about 7% compared with those obtained by addition of Pauling's values of the atomic radii for C.N.12. This suggests that there is appreciable 104.

Th-X interaction. During an investigation of the crystal (28) structures of ThAl and Th Al the present author 3 3 2 found that since the scattering powers of thorium and aluminium differ by a factor of about 8, it was not possible to determine the aluminium atomic parameters in those compounds from intensity calculations. Accordingly, the aluminium parameters suggested for ThA13 and Th3Al2 were derived from geometrical considerations. In a recent (80) paper, van Vucht has confirmed that his aluminium parameter for ThAl was also derived from geometrical considerations on the assumption that each large (Th) atom had 6 equidistant small (Al) neighbours. Since this condition is not satisfied in several other compounds with this structure, e.g. ThCo, PuNi, CeNi, DyGa, and HfA1 it may not be necessary in ThAl. A small decrease in the aluminium parameter would lead to an increase in Al-Al and a rather smaller decrease in Th-Al.

5.2.4. The compounds Th3Pd5 and Th3Pt5. Compounds with this structure do not occur in any of the other four systems under investigation and a search of the literature has failed to reveal any other compounds which have this structure. The co-ordination around the thorium atoms is 16 while that around Pd and Pd is 9 1 11 and 10 respectively. This is in general agreement with 105. (19) the suggestions of Frank and Kasper and of (20) Kripyakevich that a high co-ordination around the larger atoms might contribute to the stability of certain strucutres.

5.2.5. The compound (3-ThRh2. So far as is known, no other compounds of thorium and only one of rhodium, Sn2Rh3, have been reported to have this structure. The observed axial ratio of 1.26 and the fact that the compound occurs close to the ideal (15) composition is in agreement with Raynor's observations that the axial ratio of this type of compound tends towards 1.22 as the composition approaches the ideal one of A B. 2 0 The shortest ThRh distance in p-ThRh2 (2.67A) is considerably less than the sum of the atomic radii

(rTh rRh = 2.96) and is also shorter than that observed in any other compounds of rhodium and thorium. While this short distance suggests that there is some Th-Rh interaction it is probably due more to the geometry of the structure than to strong Th-Rh interation. 106.

5.2.6. The Laves phases (AB2). The lattice parameters obtained in the present work (43,64) show good agreement with those of Dwight and of (67) Matthias et al Dwight has pointed out that ThOs2 has a larger lattice parameter than ThIr2 although the atomic diameter of iridium is larger than that of osmium. A similar effect occurs in the isostructural compounds of uranium, lanthanum and cerium with iridium and osmium. (43) Dwight recently surveyed the contraction in the B-B interatomic distance in 168 Laves phases and found that the radius ratio fA and the position of the B element in rB the periodic table had a large effect on the B-B contraction and that the contraction in compounds with ruthenium and osmium tended to be less than that in compounds with rhodium, iridium and platinum. The compounds with thorium thus follow this general pattern.

5.2c7. Compounds of the type ThX3. 5.2.7.1. Compounds with the Cu Au (L12) structure. 3 This structure occurred in the systems with rhodium and palladium, in each case deviations from the ideal structure were observed. In the Th-Rh system, the compound existed over a narrow range of composition and the lattice was cubic at the rhodium limit and tetragonal at the thorium limit of composition. As described earlier, 107. x-ray patterns of ThRh were less sharp after the 3 material had been annealed at 1000°C than they had been in the as-melted condition. This suggests that there was greater strain in the annealed material. Possible reasons for this are a) pick-up of contaminating gases, e.g. N 2 or during annealing and b) if the lattice were cubic at 02 1000°C, the rapid cooling of the powder might prevent a strain free tetragonal lattice from forming but this would not account for the increased strain in the cubic material unless the equilibrium state at room temperature were tetragonal and the reaction rate slower at this composition.

It should be possible to obtain further information on the formation of this tetragonally distorted lattice by annealing powders from alloys of 73-77% rhodium at lower temperatures and varying the cooling rates before the x-ray examination. High temperature x-ray examination might show whether tetragonality could be a result of contamination and the use of a diffractometer might be of advantage in this approach.

(65) The lattice parameter given by Dwight for cubic ThRh was rather lower than that obtained in the present 3 work and gives a volume of the unit cell which is slightly larger than the volume of the present tetragonal cell. 108.

In the Th-Pd system this structure was observed over a narrow range of composition at approximately 78.5 - 80.5% palladium and the compound has been called ThPd4. It was not present in alloys of 77% or less palladium. The x-ray films of ThPd4 were extremely sharp indicating that lattice strain was low. Few alloy systems are known in which both an AB and an 'off-composition' 3 AB compound occur although Pells(72) has recently 3 observed that similar compounds occur in the U-Pd system. As far as is known, the Ti-Pd, Zr-Pd and Hf-Pd systems have not been examined in comparable detail and it would be interesting to know whether L12 compounds occur there. The similarity in the scattering factors of zirconium and palladium might complicate the detection of a superlattice in that system.

The contractions in the Th-Rh and Th-Pd distances in ThRh3 and ThPd4, when compared with the differences in the atomic diameters (C.N. 12) of the component metals, are close to those predicted on the basis of Dwight and Beck's(47) analysis of corresponding distances in 13 other compounds with this structure formed by elements of groups 4A and 5A with rhodium and iridium. This indicates that the bonding in ThRh3 and ThPd4 is similar to that in the other compounds in spite of the fact that ThPd4 is an 'off-composition' phase. 109.

5.2.7.2. Compounds with the TiNi 3 (DO24) structure. The structure of ThPd was determined by the 3 present author before Dwight's results were published(65) Dwight examined only one alloy and was therefore able to give only one value for the lattice parameters of this compound. In the present investigation it has been established that ThPd3 can exist over a narrow range of composition from approximately 75 - 77.5% palladium. The lattice parameters obtained on an annealed powder by Dwight (though he does not state his annealing temperature) are in reasonable agreement with those obtained at the thorium limit of composition in the present investigation. Since the range of composition over which ThPd exists is 3 on the palladium side of the ideal composition, it is suggested that this is achieved by random replacement of some thorium atoms by palladium atoms. This would also account for the observed volume contraction with increasing palladium content. In an investigation of the U-Pd system, (71) Catterall observed that UPd3, which is isostructural with ThPd3, could exist over a narrow range of composition with little variation in the "a" parameter but with the c "0" parameter varying sufficiently for 7 to be 1.67 with excess uranium present and 1.65 with excess palladium. It will be seen in section 4.3.5. that this is analogous to the variation in the lattice parameters of ThPc17. 110.

As mentioned previously (2.2.7.) Dwight and Be447) showed that in 6 compounds with this structure, the observed interatomic distances between unlike atoms were smaller than those calculated for C.N. 12 and this contraction increased with the difference in the atomic diameters of the component atoms. The observed contraction in ThPd 3 agrees well with this generalisation but the corresponding contraction in UPd , which Dwight and Beck 3 did not include in their review, is less than predicted and suggests that some other factor should be considered in the case of UPd3.

5.2.8. Compounds of the type ThX5. In his published work, Dwight does not mention the presence of extra lines on his x-ray film of ThIr5 although in a private communication he admitted that some extra unidentified lines were present. Lattice parameters for ThIr5, calculated on the basis of the hexagonal D2d structure, show excellent agreement with those given by Dwight(45 .

The pattern from material which had been annealed for one day at 1200°C was identical with that from the as- melted sample and it would be interesting to know the effect of longer anneals at this and higher temperatures on the x-ray pattern. Present indications are that the true 111. unit-cell of ThIr5 is more complex than that originally proposed and that it is closely related to the D2d structure. There is therefore some justification in discussing this compound in relation to those with the D2d structure.

ThIr was one of the 27 compounds with this struc- 5 ture surveyed by Dwight and the radius ratio of Th:Ir (1.325) is within the limits of other compounds with this structure. Dwight did not include ThNi or any compounds 5 of iron in his survey. Since ThIr is the only known 5 compound of this type, it was not possible for him to include it in his detailed study of the relationship between axial ratio and radius ratio. For other compounds, which included ThCo , it was found that for a given small 5 element, T tended to decrease as the radius ratio increased. The axial ratio of ThNi is a good fit in that respect to 5 the other compounds of nickel while the axial ratio of ThIr5 was of the same order as that of the other compounds of this type. The contraction in the observed distance between the smaller atoms as compared with that for CN 12 in the pure metals was also found to be dependent on the axial ratio and the observed Ir-Ir and Co-Co contraction in ThIr and ThCo is in good agreement with that of the 5 5 other small elements but the Ni-Ni distance in ThNi is 5 rather low. 112.

5.2.9. The types of crystal structure observed. Although the structures of many compounds in these alloy systems have still to be determined, it is now possible to make a few observations on the types of structures which are known to occur and on the interatomic distances in these compounds. The structures are characterised by a high co-ordination around the larger (Th) atoms, 15 in Th7X3 and Th2Pd, 16 in Th3Pd5 and the Laves phases, 17 in ThX and 13-ThRh2, 18 in ThRh3, ThPd3 and ThPd4 and 20 in ThIr5. It will be seen that the total co-ordination tends to increase with the percentage of smaller element in the compound.

Several of the compounds are isostructural with compounds formed between transition metals and the non- metals boron and silicon, e.g. Th2Pd with Ta2B, Co2B and Zr2Si; Th Ru etc. with Ru B • ThRu etc. with CrB, WB 7 3 7 3' and CaSi and ThRh3 with USi3. In a recent paper, (34) Aronsson has reviewed the borides and silicides of the transition metals and has pointed out that many specula- tions on the fundamental nature of the electronic interactions in these compounds have been advanced and that some of them are definitely irreconcilable with one another. One is thus not justified in making any general statements about electron transfer in these compounds although it 113. seems as if interactions between metal and non-metal may be of importance to their cohesion.

It is well known that the relative diameters of the two types of atoms are of importance in the occurrence of the CuAl2-type of structure and it is probable that this is a primary factor governing the formation of the other structures and thus explains why many "boride-like structures" occur in metal-metal systems.

It is also of interest that, apart from the Cu3Au structure which is known only in one silicide and in no borides, compounds richer in transition metal are not isostructural with borides and silicides. Aronsson pointed out that in the structures of the metal-rich boride and silicide phases, contiguous metal atoms form a three dimensional skeleton in which isolated non-metal atoms are located. With increasing non-metal content, the structures gradually become dominated by non-metal atom frameworks. In the present series of compounds, Aronsson's non-metal atom positions are occupied by the platinum-group atoms but the same geometrical considerations apply and there is a gradual transfer of dominance from the thorium to the platinum-group atom. The reason why particular structures occur only in certain of the present alloy systems is still unknown. 114.

The closest interatomic distances observed between unlike atoms in the different structures are summarised in Table XXVII and it will be seen that in the Th-Pd system the Th-Pd distance increases with increasing thorium composition of the compounds. In the other systems, the general tendency is for a decrease in the Th-X distance with an increasing amount of thorium in the compound. This implies that the mechanism of the Th-Pd bond differs from that of the Th-X bond in the other systems and this might be one reason for the different structures of the palladium compounds.

5.3. Comparison with expectations. On the basis of the simple Hume-Rothery size-factor rule, substitutional solid solutions would not be expected between thorium and the "platinum metals". From Pauling's classification of the elements into hypo-electronic, hyperelectronic and buffer atoms (see section 2.2.1.) one would expect that in the 6 alloy systems under investigation, intermetallic compounds would occur. Similarly, on a simple electrochemical approach, the difference in electronegativity between thorium and the 6 metals suggests that there is a high probability that intermetallic compounds would form. 115.

The results of the present experiments agree with these predictions as far as intermetallic compound formation is concerned. The predictions of very restricted terminal solid solutions apply to all the systems except for the solubility of thorium in palladium. Here, there is appreciable (15%) solid solubility at 1000°C in spite of the fact that the size-factor of thorium with respect to palladium is 31% (using the closest distances of approach in the unalloyed metals)(79). The behaviour of palladium will be discussed in more detail in a later section.

5,4. Comparison with other alloy systems. As mentioned earlier, our detailed knowledge of many alloy systems which involve the 'platinum-group' metals and other transition elements has still many gaps. The present comparisons will therefore be very incomplete.

5.4.1. Alloys of thorium with iron, cobalt and nickel. Intermetallic compounds occur in each of these systems and Baenziger's results(23) indicate that there are fewer compounds in the 1h-Fe system than in the other two. This is in agreement with the present results on the heavier metals of group 8. A few of the compounds are isostructural with those of the present series and 116. the similarity is greatest in the thorium-rich alloys, in each case the thorium-rich compound is of the type Th7X5 (D102). Considering the equiatomic alloys, iron like osmium does not form a compound of this composition while ThCo is isostructural with ThRu etc. and ThNi and ThPd have different structures.

In alloys containing more transition metal, the only similar structures known at the present time are ThNi and ThCo which are isostructural with Dwight's 5 5 ThIr5.

Little is known about the phase diagrams of thorium with iron and cobalt but the Th-Ni diagram was (81) given by Horn and Wasserman. As in the present series of alloys, a low melting point eutectic occurs between thorium and ThNi3 but Th7Fe3 is formed peritec- (42) tically. Some solubility of iron in thorium and of thorium in iron has been reported but since that work was performed with impure thorium, the results are of doubt- ful validity.

5.4.2. Alloys of titanium, zirconium and hafnium with group 8 metals. Although the information on many of these systems is still incomplete, inter-metallic compounds are known 117. to occur in each system. Since the sizes of the titanium, zirconium and hafnium atoms are considerably smaller than that of thorium few of the compounds are isostructural with the corresponding thorium compounds. There is a tendency to form compounds rich in group Li-A element. These cover a wider range of structures than do the thorium-rich compounds although Zr2Ni and Hf2Ni are isostructural with

Th2Pd. Considering the equiatomic alloys, the CsC1 structure is common but ZrNi and HfNi are isostructural with ThCo etc.

Zirconium like thorium, forms Laves-type compounds with ruthenium, osmium and iridium but unlike thorium, it also forms them with iron and cobalt; the Laves phases HfFe2, HfCo and HfOs have also been reported. A 2 2 Cu Au-type phase is formed by all four elements with 3 rhodium and, unlike the Th-Ir system, a similar compound is formed by titanium, zirconium and hafnium with iridium. The compounds TiPd ZrPd3, HfPd and ThPd are isostruc- 3' 3 3 tural and titanium and zirconium are known to dissolve in palladium up to at least 10% and within this concentration range they decrease the paramagnetism of palladium at approximately the same rate as thorium does. ThNi5 is hexagonal while ZrNi5 and HfNi5 have the cubic UNi5 structure but Dwight has pointed out that the occurrence of the latter structure is governed mainly by the relative sizes 118. (43) of the atoms.

The present information suggests that there is a stronger resemblance between the structure types occurring in alloys of zirconium and hafnium with those of thorium than in alloys of titanium with those of thorium. Since there are many other points of similarity between titanium and zirconium and the atomic radius of titanium is appreciably less than that of zirconium or hafnium, the similarities which do exist suggest that the type of bonding which occurs in titanium, zirconium and hafnium compounds often resembles that in thorium compounds but that where other factors, e.g. radius ratio, are dominant the thorium .structures may differ.

5.4.3. Alloys of uranium and plutonium with the group 8 metals. Intermetallic compounds are known to occur in all the uranium and in several of the plutonium systems. As with the group 4A metals, the crystal structures of the uranium-rich and plutonium-rich compounds are characteristic of uranium and plutonium compounds and are not isostructural with either the thorium compounds or with those of the other metals of group 4A although in some cases, e.g. U6Fe and

Pu6Fethe uranium and plutonium compounds are themselves isostructural. The relative sizes of the thorium atoms on 119. the one hand and the uranium and plutonium atoms on the other hand may be one reason for this difference and could also account for the solubility of the group 8A metals in 0-uranium. Several compounds with less uranium and plutonium are isostructural with the corresponding thorium compounds, e.g. URh3, UPd3, UPd4, U0s2, U1r2, PuNi5 and Pu2Ni171 while PuNi has the same structure as ThRu etc. Apart from UPd4, these uranium compounds are isostructural with the corresponding compounds of the group 4A metals.

No uranium-rich compounds occur in the U-Pd or U-Pt systems, the first compounds being UPd (which is only stable between 971° and 1047°C) and UPt.

As in the thorium systems there is a tendency to form low-melting eutectics in uranium-rich alloys with all the group 8 metals and the eutectic temperatures in alloys with iron, cobalt and nickel tend to be lower than those with elements of the two later periods.

U-Pd alloys of more than 75% palladium have many similarities with the corresponding Th-Pd alloys:- a) the axial ratios of UPd and ThPd vary in 3 3 the range 1.67-1.65. b) a compound with the Cu3Au structure occurs over a narrow range of composition at approximately 80% palladium. 120.

c) there is a minimum in the liquidus of palladium- rich alloys in the range 84-87% palladium. d) the solubility of uranium and thorium in palladium at 1000°C is approximately 15%. e) an intermetallic compound occurs with approximately 17-19% uranium or thorium and x-ray patterns of these compounds are similar although the relative positions of corresponding reflections indicates that the thorium compound has a larger unit cell.

Owing to the size difference between thorium and uranium, it is possible that electronic factors are largely responsible for these similarities.

5.4.4. Alloys of lanthanum and cerium with the group 8 metals. The information on these systems is very incomplete and nothing is known about the Ce-Pd system. Cubic Laves phases are formed by both lanthanum and cerium with ruthenium, osmium and iridium and also with nickel, rhodium and platinum and by cerium with iron and cobalt. Cerium and nickel form several compounds which are isostructural with the present thorium compounds, namely Ce7Ni3 (D102), CeNi (Bf), CeNi2 (C15) and CeNi5 (D2d). Altogether 6 compounds of lanthanum and cerium with the last structure have been reported. 121.

It is possible that many more compounds of cerium and lanthanum will be found to have the same structures as some thorium compounds since the sizes of the atoms are similar.

5.5. The behaviour of palladium in intermetallic systems. 5.5.1. General. Although there are few transition elements other than thorium with which the alloying behaviour of the 6 'platinum group' metals have been studied in detail, a few general points may be mentioned which illustrate that palladium tends to behave somewhat differently from the other elements, in some cases nickel and platinum behave similarly to palladium. a) Except for palladium, all the metals of group 8 tend to form Laves phases with scandium, yttrium, the rare earth metals, the group 4A metals, uranium and plutonium. b) Only nickel, palladium and platinum are known to form AB compounds of the DO type. 3 24 c) No XPd compounds of the D2 type are known. 5 d

Some further ways in which palladium behaves in a manner dissimilar to the other metals of group 8 may now be added:- d) Except for palladium, they all form compounds of the type Th7X3. 122.

e) Palladium is the only metal of this group to form a compound of the C16 type with thorium and this compound is isostructural with Th2Cu, Th2Ag and Th2Au. Nevitt and Downey(82) have recently pointed out that zirconium and hafnium form a series of compounds having the Si Ho (C11 2 b) structure with the elements palladium, copper, silver and gold. f) The crystal structure of ThPd differs from that of the other ThX compounds with group 8 metals. g) Palladium is the only metal of group 8 to form both an AB3 compound (D024) and an off-composition AB3 compound (L12) with thorium and uranium. h) Palladium is the only metal of group 8 to dissolve an appreciable amount of thoritim at 1000°C.

5.5.2. Thorium-palladium alloys. The solubility of thorium in palladium at 1000°C has been described (section 4.3.1.2.), since the lattice parameters of only a few unanalysed alloys were determined in this range it is not known accurately whether the lattice parameter/composition curve shows a deviation from. Vegard's law although a plot using nominal compositions suggests that any deviations are small. This indicates that, to a first approximation, the thorium atoms have the same volume in the palladium solution as they have in the pure metal. 123.

The room temperature magnetic susceptibility of palladium alloys containing nominally 1,2,5,10 and 15% (83) thorium has been measured at Nottingham University and those results indicated that additions of thorium lowered the paramagnetic susceptibility of palladium, the 15% alloy being weakly diamagnetic. This suggests that thorium dissolves in palladium with a valency of 4, i.e. with a large transfer of electrons from thorium to palladium, these electrons occupying the 0.6 holes/atom in the 4d band of palladium and consequently decreasing the paramagnetism until the 4d band of palladium is full (at 15% thorium).

(84) In their discussions of Pd-U alloys Catterall and Bates and Leach(?) have suggested that uranium behaves as if with a valency of 6 until the 0.6 holes in the 4d bard of the palladium atoms have been filled and that thereafter it behaves as if with a valency of 4. Since it seems that thorium too, has an effective valency of 4 in palladium-rich alloys it is possible that electronic factors are largely responsible for the close similarity in the behaviour of thorium and uranium in alloys of 75-83% palladium. 124.

6. Suggestions for further work. Some lines for further work are evident since the phase diagrams of none of the systems is complete, in particular, melting points above 1500°C and several crystal structures have yet to be determined. Several additional points of detail which need to be clarified have already been mentioned:- a) The temperature and crystallographic relationships between 1.- and 0- ThRh2 b) reasons for the tetragonality of ThRh 3 c) the phase relationships in Th-Pd alloys of 79-85% palladium above 1000°C d) accurate determination of the lattice parameter/ composition relationship in palladium alloys e) a study of Th-Pd alloys of 80-100% palladium below 1000°C

f) the composition of ThOsx g) the reasons for the extra lines on the x-ray patterns of ThIr 5 h) the phase relationships in Th-Ir alloys of 66-88% iridium.

Other lines of study which should yield useful information include:- a) a detailed study of palladium-rich Ti-Pd, Zr-Pd, Hf-Pd and Pu-Pd alloys to determine the solubility 125. of these elements in palladium and to see if any other off-composition L12 or low-temperature compounds occur. b) the investigation of pseudo-binary systems between isostructural compounds where size effects are a minimum, e.g. ThOs2-ThIr2, ThRu-ThRh and ThRh3-ThPd4. c) investigations of the effect of other elements on the stability of certain compounds, in some cases size effects may be kept to a minimum, e.g. additions of nickel, rhodium and platinum etc., to replace some palladium atoms inThgdand ThPd3; additions of palladium to replace some X atoms in Th X_ compounds; additions of osmium to 7 replace some iridium atoms in ThIr and ThIr ; additions 5 of palladium to replace some rhodium atoms in ThRh and

0,--ThRh2 Tl and Th Ga and d) the possible formation of Th2 2 their interatomic distances. e) an examination of thorium rich alloys in other systems where compounds of the types Th x or Th X have 7 3 2 been reported to obtain information on any eutectics which may occur and to see how far the correlations noted in section 5.1.3 may be extended. 126.

7. Conclusions. 7.1. The thorium-ruthenium system. Terminal solid solutions are restricted and 4 intermetallic compounds occur; the crystal structures of three, Th7Ru3, ThRu and ThRu 2were determined completely and the fourth compound has been tentatively called Th3Ru2. The four compounds melt congruently, melting points of Th7Ru3, 'Th,Ru and ThRu being 1412 9 2 -1-12, 1425-112 and 1462-1-12°C respectively while ThRu2 melts above 1500e.0 Five eutectics occur, that between Th and Th Ru 7 3 melts at 12621-12°C; between Th Ru and iTh Ru \ at 1388±12°C; 7 3 - 32 between Th3Ru2 and ThRu at 1388-1-12°C; between ThRu and ThRu + 0 and ruthenium 2 at 1438- 12 C and that between ThRu2 above 1500°C.

7.2. The thorium-rhodium system. No terminal solid solution was detected in this system, 7 intermetallic compounds occur and the crystal structures of Th7Rh3, ThRh, r),-ThRh and ThRh have been 2 3 determined completely (ThRh3 was given by Dwight) while those of compounds tentatively called Th Rh Th3Rh4' 3 5 and ThRh were not solved. 5-ThRh has a simple struc- 5 2 ture and in the temperature range 1200-1300°C it transforms to a more cmplex structure which was not determined. + Th Rh forms directly from the melt at 1362-12 C while 7 3 127.

ThRh and ThRh3 melt congruently above 1500°C. Th3Rh4\ and Th Rh are formed peritectically at 1487±12 and 1450±12 3 5' °C respectively while the peritectic temperatures of p-ThRh2 and 'ThRh are above 1500°C. The eutectic between thorium 5' and Th7Rh3 melts at 1237±12°C., that between Th7Rh3 and ThRh at 1312±12°C., and that between Th Rh,' and p-ThRh 3 2 at 1425±12°C. and that between /ThRh5' and rhodium at 1450±12°C.

7.3. The thorium-palladium system. Thorium dissolves in palladium up to approximately 15% at 1000°C, at this temperature 7 intermetallic compounds are stable and the crystal structures of Th2Pd, Th3Pd5, ThPd3 and ThPd4 have been determined completely (Th2Pd was determined by Ferro) while the structures of those desig- / nated ThPd , Th Pd and ThPd were not solved. Th Pd 3 4 x 2 and ThPd crystallise directly from the melt at 1162t12 and 1412±12°C. respectively while ThPd3 melts above 1500°C. Th Pd and Th Pd are formed peritectically at 1325-12 3 4 3 5 and 1387±12°C respectively. The phase relationships in alloys of 79-85% palladium were not established. Eutectics occur between thorium and Th2Pd, Th2Pd and / ThPd' and Th Pd 'and Th Pd and have melting points of 1112-12, 3 4 3 5 1137±12 and 1212±12°C respectively. There is a sharp minimum in the solidus and liquidus of the palladium solid 128. solution at approximately 12% thorium and 1125±12°C.

7.4. The thorium-osmium system. No tel. inal solid solubility was observed in this system. Three intermetallic compounds were detected, the structures of Th 0s and ThOs have been fully determined 7 3 2 (ThOs was determined by Dwight) whilst the composition and 2 structure of the third compound ThOsx are unknown. Th0s 3 crystallise directly from the melt at tempera- and ThOs2 tures above 1500°C. while ThOsx is formed peritectically and again the temperature is above 1500°C. Eutectics between thorium and Th 0s and between Th70s3 and ThOs 7 3 x melt at 1287±12 and 1482±12°C respectively while that between ThOs and osmium melts above 1500°C. 2

7.5. The thorium-iridium system. Terminal solid solubility was again restricted in this system which contains 6 intermetallic compounds. The crystal structures of Th have been 71r3, ThIr and ThIr2 fully determined (ThIr2 was given by Dwight) and Dwight has proposed a structure for ThIr5 which does not account for all the observed x-ray lines; the structures of ThIrx and 'ThIr3' are still unknown. Th7Ir3 and ThIr melt congruently above 1500°C whilst ThIrx is formed peritectically above 1500°C and decomposes eutectoidally at a temperature 129. above 1400°C. Eutectics between thorium and Th71r3 and melt at 1337±12 and 1462±12°C. between Th71rIr3 and ThIrx and between ThIr5 and while those between ThIr and ThIr2 iridium melt above 1500°C. Phase relationships and melting points in the composition range 66-88% iridium were not established.

7.6. The thorium-platinum system. There was no evidence of terminal solid solubility in this system in which 8 intermetallic compounds were detected. The crystal structures of Th7Pt3, ThPt and Th Pt were fully determined while those of ITh Pt 3 5 3 4 ' ILlaPt k/ThFt , /ThIpt and iThFt have yet to be solved. 2 3 5 ThPt, Th Pt and / ThPt crystallise directly from the melt 3 5 3 above 1500°C., Th7Pt3 is formed peritectically at 1362±12°C / 2 , ThPt4 and the peritectic temperatures of Th.)7Pt 4 ThPt and ThPt are above 1500°C. Eutectics between thorium and 5 ThPt and between / ThPt 5 and platinum melt at 1237:1-12 and 3 ' ' 1337±12°C respectively while those between ThPt and Th3Pt4 and between ThPt and ThFt melt above 1500°C. 2 3

7.7. Crystal structures of the compounds. In the present work it was established that:- a) Th7Ru3, Th7Rh3, Th70s3, Th71r3 and Th7Pt3 have the D10 structure. 2 130.

structure. b) ThRu, ThRh, ThIr and ThPt have the Bf c) Th Pd and Th Pt have a hexagonal structure of 3 5 3 5 a new type. has the C15 structure. d) ThRu2 e) (3-ThRhp has the B82 structure. f) ThPd has the DO structure. 3 24 compound. g) ThPd4 is an off-composition Ll2

The crystal structures of Th2Pd, ThOs2, ThIr2 and ThRh3 were confiEmed.

7.8. General. The results are in general agreement with alloy theory although the solubility of thorium in palladium would not be predicted by current theories. There are several points of similarity between the present systems but the constitution of Tb-Pd alloys is often different from that of the other 5 systems. The thorium-rich eutectic compositions and temperatures show a general correlation with the melting points of the component phases similar to that reported for eutectics formed by magnesium, the rare earths, uranium and plutonium. .Available results on the structures of the intermetallic compounds have been discussed in relation to other compounds with the same structures and results on the phase diagrams have been 131. compared with the available information on several other alloy systems of the group 8 metals. The structures of compounds containing less than 50% group 8 element tend to be characteristic of the other element and size factors tend to be important in determining which structures will occur. With more than 50% group 8 element, size factors are still important but the observed structures can often tolerate a much wider range of radius ratio.

Papers describing some of the results given in this thesis have appeared in the published literature. Reprints of these, together with reprints of some of the author's earlier papers are enclosed in a pocket attached to the back cover of the thesis. 132. Appendix 1. Sources and analyses of starting materials.

Thorium. Supplied by Metallurgy Division A.E.R.E. Harwell. A preliminary survey was made with material of Batch 1 prepared by electrolytic reduction and the majority of the alloys were prepared from Batch 2 which is iodide material. Batch 1. Si 30 ppm. Fe 25 ppm. Cu 20 ppm. Al 10 ppm. C 265 ppm. N 33 ppm. O .11-.13% Other elements4C5 ppm. Batch 2. Fe 20 Ppm. C 20 ppm. N 10 ppm. O 184 ppm. Other elements<5 ppm.

Ruthenium. Supplied by Johnson Matthey and Co. Ltd. in the form of sponge. Impurities detected Pd 5 ppm. Si 2 ppm. Na 1 ppm. Cu 1 ppm. Mg 1 ppm. Ag 1 ppm.

Rhodium. Supplied by Johnson Matthey and Co. Ltd. in the form of sponge. Impurities detected Ir 10 ppm. Ca 2 ppm. Fe 2 ppm. Na 2 ppm. Cu 1 ppm. Si 1 ppm.

133.

Palladium. Supplied by the Mond Nickel Co. Ltd. in the form of sponge. Impurities detected Pt 130 ppm. Rh 70 ppm. Au 70 ppm. Fe 20 ppm.

Osmium. Supplied by Johnson Matthey and Co. Ltd. in the form of sponge. Impurities detected Si 2 ppm. Na 1 ppn.

Iridium. Supplied by Johnson Matthey and Co. Ltd. in the form of sponge.

Impurities detected Si 5 mom. Fe 2 ppm. Al 1 ppm. Cu 1 ppu. Na 1 ppm.

Platinum. Supplied by Johnson Latthey and Co. Ltd. in the form of thermopure wire. Typical impurities Si 3 ppm. Fe 1 ppm. Pd 1 ppm. 134.

Acknowledcrement. This work was financed by an extra-mural research contract with the Atomic Energy Research Establishment, Harwell. The author would like to thank many friends, both at Imperial College and at the Atomic Energy Research Establishment, for much advice and many helpful discussions. Particular thanks are due to Dr. Pill. B. Waldron and Dr. M. Poole of the Metallurgy Division, A.E.R.E. for their help and co-operation in allowing the author to make use of some of their experimental facilities. Thanks are also due to Mr. C.H. Thomas of the same Division for his invaluable assistance with the melting and heat treatment of the alloys and to Mr. N. Currey of the Solid State Physics Division, A.E.R.E., whose co-operation made possible the refinement of the atomic parameters of Th3Pd5. Finally, the author would like to thank Professor J.G. Ball, under whose supervision this work was carried out, for making the work possible. 135. References 1. J.J. Katz and G.T. Seaborg, "The chemistry of the actinide elements", Methuen, 1957. 2. M.B. Waldron, J. Contemp. Phys. 1961, 2, 385. 3. J. Friedel, "Symposium on Rare Metals", Indian Inst. of Metals, 1957, 386. 4. E.P. Wohlfarth, Proc. Leeds Phil. Soc. 1948, 5, 89. 5. F.E. Hoare, J.C. Matthews and J.C. Walling, Proc. Roy. Soc. A. 1952. 216, 502. 6. J. Wucher, Ann. Phys. 1952, 7, 317. 7. L.F. Bates and S.J. Leach, Proc. Phys. Soc. B. 1956, 69, 997. 8. D. Gerstenberg, Ann. Phys. 1958, 2, 236. 9. w. Hume-Rothery and B.R. Coles, Advances in Physics, 1954, 3, 149. 10. S.L. Altmann, C.A. Coulson and W. Hume-Rothery, Proc. Roy. Soc. A. 1957, 240, 145. 11. L. Pauling, Proc. Nat. Acad. Sci. U.S.A. 1950, 36, 533. 12. K. Gschneidner and J.T. Waber, Rare Earth Symposium 1959. Am. Soc. Metals, 1960. 13. J.T. Waber and K. Gschneidner, "Plutonium 1960", Cleaver-Hume Press, London. 14. L. Pauling, J. Am. Chem. Soc. 1947, 68, 542. 15. G.V. Raynor, N.P.L. Symposium No.9, 1958, Paper 3A. 16. 0. Kubaschewski and H.A. Sloman, N.P.L. Symposium No.9, 1958, Paper 3B. 17. 0. Kubaschewski, N.P.L. Symposium No.9, 1958, Paper 3C. 18. J.S. Kasper, "Theory of alloy phases", Am. Soc. Metals Symposium, 1956. 136.

19. F.C. Frank and J.S. Kasper, Acta Cryst . 1958, 11, 184 and 1959, 12, 483. 20. P.I. Kripyakevich, Soviet Phys. Cryst. 1960, 5.2. 69. 21. K. Schubert, Z. Metallk, 1953, 44, 254. 22. K. Schubert, Z. Metallk, 1955, 46, 44. 23. J.V. Florio, N.C. Baenziger and R.E. Rundle, Acta Cryst. 1956, 9, 367. 24. R.B. Roof, A.C. Larson and D.T. Cromer, Acta Cryst. 1961, 14, 1084. 25. B. Aronsson, E. Stenberg and J. Aselius, Acta Chem. Scand. 1960, 14, 733. 26. B. Aronsson, Acta Chem. Scand. 1959, 13, 109. 27. F. Laves, "Theory of alloy phases", Am. Soc. Metals Symposium, 1956. 28. J.R. Murray, J. Inst. Metals, 1955-56, 84, 91. 29. M.E. Kitkpatrick, D.M. Bailey and J.F. Smith, Acta Cryst. 1962, 12, 252. 30. J.R. Murray, J. Less-Common Metals, 1959, 1, 314. 31. N.C. Baenziger, R.E. Rundle and A.I. Snow, Acta Cryst. 1956, 9, 93. 32. W.B. Pearson, "A handbook of lattice spacings and structures of Metals and alloys", London, Pergamon Press, 1958. 33. P.I. Kripyakevich, Soviet Phys. Cryst. 1962, 6, 501. 34. B. Aronsson, Arkiv for Kemi, 1960, 16, 379. 35. L. -E. Edshammar, Acta Chem. Scand. 1961, 12, 403. 36. C. Zener, Phys. Rev. 1951, 81, 'I'IO; 82, 403; 83, 229; 1952, 85, 324. 37. G.V. Raynor, "Progress in Metal Physics", Vol.1 London, Butterworths, 1949. 137.

38. F. Laves and H. Witte, Ketallw. 1935, 14, 645; 1936, 15, 840. 39. K.H. Lieser and H. Witte, Z. Metallk, 1952, 43, 396. 40. R.L. Berry and G.V. Raynor, Acta Cryst. 1953, 6, 178. 41. K. Kuo, Acta Het. 1953, 1, 720. 42. Rough and A.A. Bauer, USAEC. Report BM1, 1300, 1958. 43. A.E. Dwight, Trans. Am. Soc. Metals, 1961, .511 479. 44. D.I. Bardos, K.P. Gupta and P.A. Beck, Trans. Met. Soc. AIME, 1961, 221, 1087. 45. G.B. Brook, G.I. Williams and E.M. Smith, J. Inst. Metals, 1954-55, 83, 271. 46. J.H. Wernick, S.E. Haszlo and D. Dorsi, J. Phys. Chem. Solids, 1962, 23, 567. 47. A.E. Dwight and P.A. Beck, Trans. Het. Soc. AIME, 1959, 215, 976. 48. K. Schubert, N.P.L. Symposium No.9, 1958, Paper 3D. 49. A.E. Dwight, Private communication. 50. W. Hume-Rothery, G.W. Mabbott and K.E. Channell- Evans, Phil. Trans. A. 1934, 233, 1. 51. T.B. Massalski, "Theory of alloy phases", Am. Soc. Metals Symrc.sium, 1956. 52. A.G. Knapton, J. Savill and R. Siddall, J. Less- Common Metals, 1960, 2, 357. 53. R.J. Douglass and E.F. Adkins, Trans. Met. Soc. AIME, 1961, 221, 249. 54. A.R. Powell, Pt. Metals Rev. 1960, 4, 144. 55. E. Raub, J. Less-Common Metals, 1959, 1, 3. 56. A.A. Rudnitskii and 0.A. Novikova, Russ. J. Inorg. Chem. 1959, 4, 719 and 722. 138.

57. A.A. Rudnitskii and A.N. Khotinskaya, Russ. J.. Inorg, Chem. 1959, 4, 1053 and 1160. 58. A.A. Rudnitskii and V.P. Polyakova, Russ. J. Inorg. Chem. 1959, 4, 1051 and 1158. 59. J.R. Murray, UKAEA Report .ERE M/R 2242, 1958. 60. D.S. Evans, G.V. Raynor and R.T. Weiner, J. Nucl. Mat. 1960, 2, 121. 61. D.S. Evans, Birmingham University, Progress Report 1960. 62. D.S. Evans and G.V. Raynor, J. Less-Common Metals, 1961, 3, 179. 63. G.H. Bannister and J.R. Thomson, UKAEA Report in preparation. 64. A.B. Dwight, J.W. Downey and R.A. Conner Jr., Trans. Met. Soc. AIME, 1958, 21.21 337. 65. A.E. Dwight, J.W. Downey and R.A. Connor Jr., Acta Cryst. 1961, 14, 75. 66. R. Ferro and G. Rambaldi, Acta Cryst. 1961, 14, 1094. 67. B.T. Matthias, V.B. Compton and E. Corenzwit, J. Phys. Chem. Solids, 1961, 19, 130. 68. R. Ferro and R. Capelli, Acta Cryst. 1961, 14, 1095. 69. J.R. Murray and G.K. Williamson, J. Less-Common Metals, 1959, 1, 73. 70. S.J. Lloyd and J.R. Murray, J. Sci. Inst. 1958, 15, 252. 71. J.A. Catterall, J.D. Grogan and R.J. Pleasance, J. Inst. Metals, 1956-57, 85, 63. 72. G.P. Pells, AERE, Harwell, Private communication. 73. G.V. Raynor, "The physical metallurgy of magnesium and its alloys", Pergamon Press, 1959. 74. J. Betterton and J. Frye, Acta Met. 1958, 6, 205. 75. J.R. Murray, J. Inst. Metals, 1958-59, 349. 139.

76. G.H. Bannister, R.C. Burnett and J.R. Murray, J. Nuc. Materials, 1960, 2, 51. 77. R.M. Goldhoff, H.R. Ogden and R.I. Jaffee, US EC Report BMI 720, 1951. 78. D.S. Evans and G.V. Raynor, J. Nuc. Materials, 1959, 1, 281. 79. L. Pauling, "Theory of alloy phases", tm. Soc. Metals Symposium, 1956. 80. J.H.N. van Vucht, Philips Research Reports, 1961, 14, 1. 81. L. Horn and C. Wasserman, Z. Metallk, 1948, 39, 273. 82. M.V. Nevitt and J.W. Downey, Trans. Met. Soc. AIME, 1962, 224, 195. 83. Nottingham University, Physics Department, Unpublished results, 1962. 84. J.A. Catterall, Phil Mag. 1957, 2, 449.

Hypoelectronic Stable atoms Hyperelectronic atoms atoms Cl a ssi Li Be B C N 0 F f i el c ati

Na Mg Al Si P. S Cl e ct o n ro n t Buffer atoms of

ra at f-9 nsf K Ca Sc Ti V Cr* Mn Fe Co Ni Cu Zn Ga Ge As. Se Br o ms td H

e 8 wi Rb Sr Y Zr Nb Mo* Tc Ru Rh Pd Ag Cd In Sn Sb Te I r pro t I H h pe res Cs Ba La Hf Ta W4 Re Os Ir Pt Au Hg Tl Pb Bi Po At rti pect t es . o th These atoms can accept electrons but cannot give electrons without change in valence. ei r

141. TABTF, II Comparison of observed and calculated sin28 values and line intensities for Th7Ru3 _f ... 2_., 2_.._..i .....7.______...... , 2 2 hkl sin 0 sin 26 I 1 hkl 1sin e sin 0! I i I obs. calc. obs. calc t - obs ' calc.'obs.lcalc . 100 .0079 .0080 VVW 2 1 312 I.1633 .1632; VVW 1 4 001 .0149 0 321 .1664 !( .166211\2S (133 101 . .0229 .0228 VW , 27 ' 203 ;(.1663, ( 4 110 ;.0239 .0239 VVN 15 '1 410 .1670 1 .1672! VVW 9 200 .0318 0.7 '1 411 .1820 .1821:M 109 111 1 .0388 0 402 .1868 .1871 'IT 42 201 .0467 .0468 W 46 213 .1901 1 .1902i MW 74 210 .0557 .0557 MS 236 500 .1991 .1990 : W 38 002 .0597 .0598 VVW 5 303 .20 60 ! . 2061 MS 171 102 .0676 .0677 VS 747 322 .2109 .2110 MS i 282 211 .0705 .0707 VVS 1000 501 .2139 .2139'MS 269 300 .0716 .0716 M 165 .2151: . 2149 MS 203 .0835 ! 3 112 i .0836 MW 87 L01-2 30 .2224 .2229 VVW 2 301 .0864 .0866 VS 596 412 .2270 .2269W 55 202 .0915 .0916 S 514 331 - . 2299 - 0 1, 220 .0955 .0955 S 386 223 - .2300i - 0 310 .1037 .1035 VVW 421 .2380 . 2378! MW ( 15

221 .1105 0 313 i(. 2379) .( 72 212 .1155 .1155 MW 74 004 .2389 ! .2390 W 43 311 .1184 .1184 MW 83 510 .2467 (.2467 s VW ( 1 400 .1275 .1274 VW 27 104 !(.2470 ( 17

302 .1314 .1314 W 51 a 502; .2587 .25881 VW 33 11 003 - .1345 0 11 511 .2616 (.2617 MW ( 45 401 .1423 (.1423 MW (76 403 (.2618 ( 43 103 (.1424 (19 114 .2629 - 320 .1513 j .1512 VW* 3 204 .2708 i .2709 VVW1 2 222 .1552 .1553 VVW 3 332i .2746 ! .27471 VW 1 21 113 , .1583 0 422 .2824 .2826 VW 24 1 L 142. TABLR II cont:

hkl sin26 sin20 [ I I hkl sin20 sin26 I I obs. calc. obs. calc obs. calc. obs. calc 323 .2857 .2857 MW 75 611 .3573 (.3572 VW (25 600 - .2866 - 0 1 423 (.3573 ( 1 430 .2946 (.2945 :W ( 3 404 .3665 .3664 VVW 3 214 1(.2948 ( 42 522 .3702 .3702 VW 22 601 .3016 K.3015 M ( 48 005 - .3735 - 0 413 k.3016 ( 51 513 .3814 (.3812 MW (59 ! 512 .3064i .3065 It 65 105 (.3815 ( 4 431 .3093;•30951 MW 66 440 - .3821 - 0.7 520 .3103 (.3104 W (0.5 530 .3905 (.3900 W ( 6 304 (.3107 ( 52 700 (.3900 (. 0 521 .3255 .3254 M 126. 324 (.3903 (47 503 .3335 j .3335 !IN 42. 441 - .3970 - 0 224 .3343 .3346 MW 951 115 - .3974 - 0 610 .3421 .3423 VVW 6 612 - .4020 - 2 314 - .3425 - 3! 531 .4051 (.4050 MW ( 4 602 - 1 .3463 - 2 701 (.4050 460 333 - 1 .3494 - 01 205 (.4053 , (22 432 .35/M 1 .3543 MW 63

i ,_. L i

VVS very very strong M medium VS very strong MW medium weak S strong W weak MS medium strong VW very weak VVW extremely weak.

Coincident with a Th02 line. 143. TABLE III X compounds Interatomic distances in Th7 3 Ru Rh Os Ir Pt 3 x 3.03 3.04 3.04 3.05 3.07 3 Th111 3.63 3.65 3.65 3.67 3.69 3.7b 3 Thill 3.66 3.67 3.67 3.68 3 Th11 3.72 3.74 3.74 3.76 3.77 3 Th111 3.98 3.98 3.98 3.98 4.01 2.90 Th11 - 2 X 2.85 2.87 2.87 2.88 1 X 2.95 2.95 2.95 2.95 2.98 2 Th111 3.59 3.60 3.61 3.62 3.64 1 X 3.67 3.67 3.67 3.67 3.70 1 Thl 3.72 3.74 3.74 3.76 3.77 2 Th11 3.77 3.78 3.78 3.81 3.83 4 Th11 3.83 3.83 3.83 3.84 3.87 2 Th111 3.83 3.85 3.85 3.87 3.89 Th111 2 X 2.98 2.98 2.98 2.99 3.00 3.1111 2 Th111 3.44 3.44 3.44 3.47 2X 3.47 3.49 3.49 3.51 3.52 2 Thll 3.59 3.60 3.61 3.62 3.64 1 Thi 3.63 3.65 3.65 3.67 3.69 1 Thi 3.66 3.67 3.67 3.68 3.70 2 Th111 3.79 3.81 3.81 3.83 3.85 2 Thll 3.83 3.85 3.85 3.87 3.89 1 Thi 3.98 3.98 3.98 3.98 4.01

X - 2 Thll 2.85 2.87 2.87 2.88 2.90 1 Thll 2.95 2.95 2.95 2.95 2.98 2 Th111 2.98 2.98 2.98 2.99 3.00 1 Thi 3.03 3.04 3.04 3.05 3.07 2 Thill 3.47 3.49 3.49 3.51 3.52 1 Thll 3.67 3.67 3.67 3.67 3.70 TABLE IV X-ray data for Th7X3 compounds

tf fl c/a Volume of x-ray Compound "a" c unit cell density 0 0 0.A A A A' gm/cc -.003+ ±.002

Th7Ru3 9.969 6.302 0.632 542.4 11.82

Th Rh 10.031 6.287 0.627 548.0 11.71 7 3 Th70s3 10.031 6.296 0.628 548.7 13.27

Th7Ir3 10.076 6.296 0.625 553.7 13.16

Th Pt 10.126 6.346 0.627 563.5 13.01 7 3 14-5.

TABLE V 2 Comparison of observed and calculated sin e values for Th7Rh3' Th70s3 , Th7Ir3 and Th7Pt3 · hk1 Th Os Ir Th pt3 Th7Rh3 7 3 Il'h7 3 7 . 2e 2 2 . 2e . 28 . 2e . 2e 2 s~n sin e sin e s~n s~n sJ..n sJ..n sin e obs. calc. obs. calc. obs. calc. obs. calc.

.~ 100 .0079 .0079 .0078 .0077 101 .0229 .0228 .0228 .0227 .0228 .0224- .0224- 110 .0236 .0235 .0236 .0233 .0234- .0231 .0231 200 .0314- .0315 .0314- .0311 .0312 .0308 .0309 201 .04-63 .04-64- .04-64- .04-61 .04-56 210 .0550 .0550 .0550 .0550 .0545 .0545 .0540 .054-0 002 .0600 .0600 .0597 .0599 .0597 .0599 .0589 .0589 102 .0678 .0679 .0677 .0677 .0676 .0677 .0665 .0666 211 .0700 .0700 .0699 .0700 .0694- .0695 .0687 .0687 300 .0707 .0707 .0707 .0707 .0700 .0701 .0694- .0694- 112 .0836 .0836 .0834- .0835 .0832 .0832 .0820 .0821 301 .0858 .0857 .0857 .0857 .0851 .0851 .0841 .084-2

202 .0915 .0915 .0913 .0913 .0910 ~0910 .0897 .0898 220 .0944- .094-3 .0944- .0943 .0935 .0935 .0926 .0926 310 .1022 .1022 .1013 .1003 212 .1151 .1151 .1150 .114-9 .1144- .1144- .1129 .1129 311 .1172 .1172 .1172 .1171 .1164- .1162 .1151 .1150 400 .1257 .1258 .1258 .1258 .124-8 .1246 .1233 .1234- 302 .1308 .1308 .1307 .1306 .1300 .1299 .1285 .1283 401 .1407 .1408 .1407 .1407 .1397 .1396 .1382 .1382

./.. 146.

TABLE V cant: hkl Th Os Th7Rh3 7 3 Th7Ir3 Th7Pt3 . 2e 2 2 2 2 2 . 26 2 s~n sin e sin e sin e sin e sin e S:Ln sin e obs. calc. obs. calc. cbs. calc. abs. calc. ------.. -,...... ~ 103 .1429 .1425 ~1426 .14-24 .1425 .1403 .1403 320 .1493 .1493 .1480 .1466 222 .1544- .1545 .1542 .1537 .1534- .1515 .1515 312 .1622 .1621 .1611 .1592 321 .1644- .1643 .164-3 .1643 .1632 ( .1630 .1614- .1613 410 .1651 .1651 L1636 .1620 203 .1664 .1665 .1661 .1662 .1659 .1659 .1634 .1634 411 .1802 .1801 .1800 .1800 .1787 .1786 .1768 .1767

402 .1855 .1858 .1856 .1856 .18L~6 .1845 .1824- .1824- 213 .1902 .1901 .1897 .1897 .1893 .1893 .1866 .1866 500 .1965 .1965 .1965 .1947 .1929 303 .2059 .2058 .2055 .2055 .2048 .2048 .2021 .2020 322 .2094- .2094- .2092 .2092 .2080 .2079 .2056 .2055 501 .2115 .2115 .2116 .2115 .2096 .2097 .2075 .2076 330 .2122 .2122 .2121 .2122 .2103 .2103 .2083 .2083 420 .2201 .2201 .2181 .2160 412 .2251 .2251 .2249 .2249 .2234 .2235 .2208 .2209 421 .2351 .2351 .2350 .2333 .2331 .2306 .2307 313 ,,2371 .2373 .2372 .2369 .2361 .2360 .2327 .2329 004 .2402 .2402 .2393 .2395 .2393 .2395 .2355 .2357 510 .2434- .2/+34- .2415 .2392 104 .2479 .2480 .2475 .2474 .2470 .2473 .2435 .2434 502 .2563 .2565 .2564 .2564 .2543 .2546 .2518 .2518

147.

TABLE VI

Comparison of observed and calculated sin20 values and line intensities for ThRu.

hkl sin ge singe obs. calc. obs. calc.

020 - .0186 - 0 110 .0439 .0441 VW 59 021 .0543 .0544 M 243 040 .0742 .07'1 II S 402 111 .0799 .0799 VS 1000 130 .0811 .0813 M 318 041 .1100 .1102 VW 159 131 .1171 .1171 S 484 002 .1432 .1432 W 231 150 - .1557 - 2 200 .1575 .1578 W 201 022 - .1618 - 0 060 - .1674 - 1 220 - .1764 - 0 112 - .1873 - 20 151 .1916 .1915 W 189 061 .2031 .2032 VW 70 221 .2122 .2122 VW 80 042 .2173 .2176 VW 167

•/ • • 148.

TABLR VI cont: hkl sin2 0 sin2 6 obs. calc. obs. calc.

132 .2244 .2245 VW 142 240 .2321 .2322 VW 148 170 .2675 (.2673 W (143 ( 241 (.2680 84 080 - .2976 - 7 152 - .2989 - 9 202 .3011 .3010 W 140 171 - .3031 - 26

VS = very strong S = strong M = medium W = weak VW = very weak ThRu ThX Thlt ThIr ThCo Th_1.1 ThRh 4.42 3.878 3.900 3.894 3.866 3.74 a A 0 11.24 11.29 10.88 11.09 11.13 11.45 b A 0 4.220 4.071 4.16 4.19 4.454 4.266 A c 0 0.136 0.140 0.1/1 0.140 0.140 0.147 YTh 1 1 0.410 0.443 0.410 0.410 0.416 0.410 Y x of unit 178.3 169.3 183.4 192.6 184.9 Volume 212.1 cell A 0 3

Density 14.72 12.24 12./1/1 11.41 15.26 gm/cc. 8.11 diameter Atomic C.N.12 of x 2.76 2.70 2.68 2.67 2.50 2.86

seangonags gao aqq. BuTAuti

ral ral u ruct t S a f a at d r T r o hX hX s s pound m co

IIA aagITI 150. TLBLE VIII Comparison of observed and calculated sin20 values and line intensities for ThRh.

hkl sin2 0 sin2 0 obs calc. obs. calc. 020 - .0188 - <. 1 110 - .0444 - 51 021 .0518 .0521 MW 237 040 .0751 .0752 M 384 111 .0777 .0777 S 1000 130 .0819 .0820 W 299 041 .1084 .1085 W 160 131 .1154 .1153 M 492 002 - .1520 - <1 150 - .1572 - 3 200 .1589 .1588 W 191 060 - .1692 - < 1 220 - .1776 - -<, a. 112 - .1776 - 19 151 .1903 .1905 W 175 061 - .2025 - 67 042 .2084 .2084 W 175 221 - .2109 - 74 132 .2152 .2152 W 149 240 .2341 .2340 W 144 241 - .2673 - 82 170 .2698 .2700 W 134 152 - .2904 - 17 202 .2922 .2920 W 145

S = strong M = medium

MW = medium weak W = weak 151. TABLE IX

Comparison of observed and calculated singe values and line intensities for ThIr.

hkl sin20 sin20 obs. calc. obs. calc. 020 .0191 .0192 VVW 24 110 - .0439 - 6 021 .0517 .0518 VVW 27 111 .0766 (.0765 VS (1000 ( ( 040 (.0766 ( 365 130 .0822 .0822 MW 180 041 .1092 .1092 M 221 131 .1149 .1148 S 659 002 .1304 .1304 M 280 022 - .1496 - 4 200 .1565 .1564 M 215 150 .1591 .1588 VVW 29 060 - .1724 - 4 112 - .1743 - 1 220 - .1755 - 3 151 .1912 .1914 W 87 061 - .2050 - 35 042 .2071 .2070 MW 173 221 - .2081 - 8 132 .2124 .2126 W 92

152.

TABLE IX cont:

hkl sin2 e sin2 8 obs. calc. obs. calc. 240 .2331 .2330 MW 143 241 .2659 .2656 W 113 170 .2741 .2738 W 128 202 .2869 .2868 M 168 152 - .2892 - 21 062 - .3028 - 3 222 .3059 2 171 .3064 .3064 VW 47 080 .3066 ( 10

VS = very strong S = strong

M = medium MW = medium weak

W = weak VW = very weak VVW = extremely weak 153.

TABLE X

Oomparison of observed and calculated sin2e values and line intensities for Thpt.

2 hkl Sln. 20 sin e I I abs. calc. obs. calc. 020 .0193 00193 VVW 24- 110 .04-38 5 021 .04-92 .0492 VVW 21 III .0736 .0737 VS 1000 04-0 .0769 .0771 MS 340 130 .0823 .0824 MW 168 04-1 .1068 .1070 MW 205 131 .1122 .1123 S 644- 002 .1196 .1196 M 295 022 .1389 4 200 .1558 .1560 MW 202 150 .1594- .1595 VVW 28 112 .1634- 1 060 .1732 4 220 .1753 :3 151 .1895 .1894- vvr 83 042 .1967 .1967 I'ifvV 179 132 .2018 .2019 W 92 061 .2033 .2034 VVW 32 221 .2052 7

./0.

154.

TABLE X cont:

2 hkl sin 0 sin2 6 obs. calc. obs. calc.

240 .2331 .2331 W 139 241 .2630 .2630 W 112

170 .2756 (.2752 114 p_20 202 .2756 (167 152 - .2791 23 023 - .2884 2 062 - .2931 4 222 - .2949 2 171 .3050 .3051 VVW 47

VS = very strong strong MS = medium strong medium MW = medium weak weak VW = very weak VVW = extremely weak. 155.

TABLE XI

Interatomic distances in ThX compounds

Co Ru Sth- Ir Pt Al Th - 4s 2.86 2.87 2.92 2.94 5.01 3.22 1 X 3.04 5.05 3.03 3.00 2.99 3.39 2 X 3.04 3.24 3.23 3.22 3.21 3.22 4 Th 3.62 5.75 5.78 3.79 3.84 .85 2 Th 3.62 3.76 3.79 3.82 3.82 5.96 2 Th 3.71 3.88 ,.87 3.89 5.90 4.19

9 2 X .71 2.77 2.86 2.92 2.93 2.99 2.46 4 Th 2.86 2.87 2.92 2.94 3.01 3.22 1 Th 3.04 3.05 3.03 3.00 2.99 3.39 2 Th 3.04 3.24 3.23 3.22 3.21 3.22

156.

TABLE XII

Comparison of observed and calculated sin 2e values and line intensities for ThRu2.

hkl sin2 8 sin 28 obs. caic. obs. calc. 111 .0303 .0304 Tyr 77 220 .0809 .0810 S 448 311 .1114 .1113 VS 1000 222 .1215 .1214 M 183 400 - .1619 - 13 331 .1923 .1923 VW 24 422 .2427 .2429 MS 235 511 .2731 .2732 S F52 333 .2732 ( 84 440 .3237 .3238 MS 258 531 .3541 .3542 VW 20 620 .4049 .4048 MW 109 533 .4352 .4352 MW 125 622 ./I)155 .4453 W 68

VS = very strong S = strong MS = medium strong M = medium MW = medium weak W = weak VW = very weak

157.

TABLE XIII

Comparison of observed and calculated sin2 0 values and line intensities for Th0s2 and ThIr2.

ThOs 2 ThIr2 hkl sin20 sin26 I I sin20 sin2e I I obs. talc. obs. talc. obs. calc. obs. talc. 111 .0297 .0299 VW 12 .0302 .0303 VW 15 220 .0796 .0797 S 329 .0808 .0809 S 327 311 .1095 .1096 VS 1000 .1112 .1112 VS 1000 222 .1196 .1196 S 349 .1213 .1213 S 352 400 .1595 .1595 VW 27 .1616 .1617 VW 28 331 - .1894 - 2 - .1921 - 3 422 .2389 .2392 M 132 .2427 .2426 M 133 511 .2688 (.2691 S (268 .2727 (.2729 S (263 ( 333 (.2691 89 L2729 87 440 .3188 .3189 MS 293 .3235 .3235 MS 285 531 - .3488 - 2 - .3538 - 2

620 .3990 .3987 W 61 .4045 .4044 71 64 533 .4288 .4286 M 131 .4349 .4347 M 134 622 .4388 .4386 M 149 .1111119 .4448 M 154 4144 .4786 .4784 VVW 7

VS = very strong S strong MS - medium strong

M = medium W = weak VW = very weak VVW = extremely weak. Not reported by Dwight (64) 158.

TABLE XIV

Interatomic distances in ThX 2 compounds

Ru Os Ir

Th - 12 X 3.18 3.20 3.18 4 Th 3.31 3.34 3.32

X - 6 X 2.70 2.72 2.70 6 Th 3.18 3.20 3.18

159. TABLE XV

Comparison of observed and calculated 2 sin values and line intensities for p-ThRh2.

hkl sin ge sin2 obs. calc. obs. calc. 100 .0369 .0369 VW 90 101 .0541 .0543 M 280 002 .0691 .0694 if 168 102 .1061 .1063 S 847 110 .1108 .1108 S 1000 200 - .1477 - 16 201 .1649 .1650 Vii 70 112 .1801 .1801 M 273 103 .1929 .1930 VW 55 202 .2172 .2170 M 288 210 - .2584 - 14 211 - .2758 - 65 004 .2775 .2771+ VW 81 203 - .3037 - 28 104 - .31/11 - 11 212 .3280 .3278 M 300 300 .3323 .3323 W 186 114 .3884 .3882 M 288

S = strong M = medium

W = weak VW = very weak 160.

TABLE XVI

Comparison of observed and calculated 2 . sin a values and line intensities for cubic ThRh3

hkl sin 0 obs. calc. obs. calc.

100 .0343 .0343 W 15 110 .0684 .0687 W 12 ill .1030 .1030 VS 1000 200 .1375 .1374 S 520 210 .1716 .1717 vi 7.5 211 .2059 .2060 VW 5.6 220 .27611 .2747 M 337 300 .3093 (.3091 vvi .--.' 1 221 .3091 3.0 310 .3435 .3434 VW 2.7 311 .3775 .3777 M 399

VS = very strong S = strong M = medium W = weak VW = very weak

161. TABTR XVII 2 Comparison of observed and calculated sin 8 values and line intensities for tetragonal ThRh3. 2 hkl sin28 sin 0 obs. calc. obs. calc. 100 .03/11 .0345 W 10 001 .0349 .0351 lw 5 110 .0693 .0690 W 4 101 .0698 .0696 ',14. 8 111 .1040 .1041 S 1000 200 .1382 .1380 M 352 002 .1405 .1404 M 170 210 .1729 ( .1725 VW ( 2 201 ( .1731 ( 2 102 .1749 .1749 VW 3 211 .2076 .2076 VW 4 112 - .2094 - 2 220 .2758 .2760 M 112 202 .2783 .2784 M 223 300 - .3105 - < 1 221 - .3111 - 1 212 .3132 .3129 VW 2 003 - .3159 - < 1 310 - .3450 - <1 301 - .3456 - <1 103 - .3504 - <1 311 .3803 .3801 M 266 113 .3849 .3849 M 133

S = strong M = medium W = weak 117 = very .weak. 162. TABLE XVIII

Phases identified in Pd-Th alloys

Annealing temperature Thorium As-melted 800 1000 1150 1200 15 Pd

17 ThPdx ThPdx ThPdx 19 ThPdx 20 ThPd + ThPd ThPd ThPd 3 4 4 4 21 - - ThPd4 - ThPd + ThPd4 22 - 3 23 ThPd - ThPd 3 3

TABTR XIX

Hardness of Pd-Th alloys after homogenising for 4 days at 1000°C.

% Thorium VPN % Thorium VPN 0 49 16 510 5 190 17 547 10 260 18 526 11 296 19 520 13 356 20 488 14 420 21 450 15 440 163

TABLE XX

Comparison of observed and calculated 2 sin values and line intensities for Th2Pd.

hkl sin 2 sin2 0 obs. calc. obs. calc. 110 - .0222 - 5 200 .0/1113 .0/03 VVW 27 002 .0669 .0668 MW 173 211 .0722 .0722 VS 1000 220 .0887 .0889 M ( 65 ( 112 .0890 ( 232 310 .1113 ( .1111 S ( 368 ( ( 202 ( .1112 ( 252 222 - .1557 - 2 321 - .1611 - 17 400 .1780 .1778 W 31 312 .1779 61 330 .2001 .2000 W 71 411 .2058 .2056 M 180 213 .2058 282 420 .2222 .2222 VW 54 402 .2447 .2446 MW 158 332 .2671 ( .2668 M ( 241 ( 004 .2672 ( 63 510 - .2887 - 1

•-/• • 164.

TABLE XX cont: hkl sin2 0 sin2 8 obs. calc. obs. calc. 422 --'.2890 - . 1 114. - .2894 - < 1 431 - .2944 - 4 323 - .2947 - 7 204 - .3116 - 4 521 .3391 ( .3389 M ( 175 ( ( 413 ( .3392 ( 83 /MO ( .3555 ( 23 ( 512 .3561 ( .3557 15 28 224 .3561 16 530 .3780 ( .3777 MW ( 22 ( 314 .3783 ( 116 600 .4005 .4000 W 61

VS = very strong S = strong M = medium MW = medium weak W = weak VW = very weak VVW = extremely weak. 165. TABTR

Comparison of observed and calculated 2 Pd . sin 9 values and line intensities for Th3 5

hklsi. n2 0si. n 28 obs. calc. obs. calc. 100 - .0155 - < 1 001 - .0390 - 8 110 .0464 .0464 S 252 101 .0543 .0545 VVW 14 200 .0617 .0619 VVW 13 111 .0854 .0855 VS 1000 201 .1009 .1009 W 77 210 .1083 .1084 M 144 300 .1394 .1393 S 226 211 .1475 .1474 MW 90 002 .1561 .1561 M 124 102 - .1716 - < 1 301 .1783 .1783 MW 83 220 .1857 .1858 M 108 310 - .2012 - 10 112 .2025 .2025 MW 67 202 - .2180 - 3 221 .2247 .2248 M 94 311 .2404 .2403 VVW 22 400 - .2477 - •-,1 212 .2645 .2644 W 68 166.

TABLE XXI cont: 2 hkl sin 0 sin2 e obs. caic. obs. calc. 401 - .2867 - 16 320 - .2941 - 7 302 .2953 .2954 M 131 410 .3254 .3251 VW 50 321 .3334 .3333 VW 30 222 .3418 .3418 W 77 003 - .3512 - <1 312 - .3573 - 7 411 .3641 .3641 MS 190 103 - .3667 - < 1 500 - .3870 - 8 113 .3976 .3976 MW 85 402 - .4038 - <1 203 - .4131 - 8 330 .4182 .4180 VVW 27

VS = very strong S = strong MS = medium strong M = medium MW = medium weak W = weak VW = very weak VVW = extremely weak. 167.

TABLE XXII

Comparison of observed and calculated sin2e values for Th Pt " 3 5

sin2e Sln. 2e sin2e sin2e hkl obs. calc. hkl oDs • calc.

100 • 0154 112 .2016 .2017 001 .0388 .0388 202 .2171 110 .0462 .0463 221 .2238 .2239 101 .0543 .0543 311 .2392 .2393 200 .0616 .0617 400 .2467 III .0851 .0851 212 .2634- .2633 201 .1005 .1005 401 .2856 210 .1080 .1079 320 .2390 300 .1388 .1388 410 .3237 .3238 211 .1469 .1468 321 .3319 .3318 002 .1556 .1554- 302 .3330 102 .1708 222 .3403 .3404 301 .1776 003 .3496 220 .1850 .1850 411 .3627 .3627 310 .2003 .2005 168. TABLE XXIII

Comparison of observed and calculated sin20 values (75% palladium) and line intensities for ThPd3 sin2 0 sin2 6 hkl obs. calc. obs. calc. * 100 .0231 .0230 VVW 19 002 - .0246 - 0 101* .0291 .0292 MW 83 102 .0477 .0477 M 139 110 .0690 .0691 MW 81 1, 103 .0786 .0785 VW 25 200 .0922 .0922 MW 76 112 - .0938 - 0 201 .0983 '.0984 S (417 ( 004 .0986 (370 202 .1169 .1168 VS 1000 104 - .1216 - 5 203 .1478 .1476 MS 237 210 - .1613 - 3 211 .1677 (.1675 MW ( 18 ( 114 (.1677 ( 51 212 .1859 .1860 W 50 204 .1908 .1908 W 53 , 300 .2073 .2074 VVW 18 301 - .2136 - 0 213 .2168 .2168 VVW 13 169. TABTR XXIII cont:

sin2 0 sin 26 hkl obs. calc. obs. calc.

006 - .2218 - 0 302 - .2320 - 0 106 - .211/18 - 16 205 .2459 .2462 MW 108 214 - .2597 - 3 303 - .2629 - 0 220 .2765 .2766 MS 239 116 - .2909 - 0 310 - .2996 - 10 222 - .3012 - 0 311 .3059 .3058 VW 8 304 .3060 21 206 .3142 .3140 MS 215 215 - .3153 - 7 312 .3240 .3243 VVW 21 107 - .3249 - 4

*Not reported by Dwight(65) VS = very strong, S = strong, MS = medium strong, M = medium, MW = medium weak, W = weak, VW = very weak, VVW = extremely weak.

170.

TABLE XXIV

Comparison of observed and calculated singe values and line intensities for ThPd4 (80% palladium)

sin ge sin2 0 hkl obs. calc. obs. calc. 100 .0350 .0351 W 104 110 .0701 .0702 W 83 111 .1055 .1054 VS 1000 200 .1405 .1405 S 516 210 .1756 .1756 W 50 211 .2108 .2107 W 37 220 .2809 .2810 MS 338 300 .3163 (.3161 VW ( 21 ( 221 (.3161 5 310 .3513 .3512 VW 18 311 .3864 .3863 MS 397 222 .4213 .4214 M 119

VS = very strong, S = strong, MS = medium strong M = medium, W = weak, 11W = very weak.

171.

TABLE XXV

Comparison of sin26 values for alloys of 66-83% iridium.

66% Ir 72°o Ir 7590 Ir 80 Ir 83°o Ir i sin26 I sin26 I sin26 I sin26 I sin26] I obs. obs. obs. obs. obs. obs. obs. obs. obs.; obs.

.0277' W I .0279 W .02791 W , .0283 VW .0286 M .0302 VW .0303 VW .0601 W .0601 M ; .0696 1M .0693 S .0699 W .0808 S .0809 S .0826 M ) .0821 S .0827 W .0826 W ) ) .0832 VS .0834 VS .0837 M .0941 VVW , .0938 W .0979 VW .0979 WM .1108 MS .1107 MS .1112 VS .1114 VS .1118 S .1116 S ! .1120 M .1137 S .1138 VS .1142 S .1151 MS 1I .1151 S .1161 S .1162 VS .1213 S .1215 S .1220 MS .1215 VW 1 .1218 VW .1239 VW .12/111 M .124x W 1 .1300 M II, .1300 S 172.

TABU', XXV cont:

66% Ir 72% Ir ' 75% Ir 80% Ir 83% Ir 2 2 2 sin 8 I sin 8 I I sin 0 I sin2 e I sin2 8 I obs. obs. obs. obs. I obs. ! nips. obs. obs. obs. obs. :J.- .1316 MS ) 1 .1309 M .1320 1 S .1328 W .1431 W .1432 S .1445 VW .1/1115 M .1524 W .1525 1 M .1533 W ) I .1539 VW .1577 VVW .1616 VW .1617 VW : .1628 VW .1941 VW .1943 VW .1945 VTN .2133 VW .2136 W .2222 VVW .2255 VVW 1 .2277 VVW .2349 VW .2356 W .2407 VW .2405 W .2427 M .2427 M

VS = very strong, S = strong, MS = medium strong, M = medium MW = medium weak, W = weak, VW = very weak VVW = extremely weak.

173.

TABLE XXVI

Comparison of observed and calculated 2 sin-8 values and line intensities for ThIr5.

sin228 averagesin 2 0 I I hkl obs. sin. 2 0 calc, obs. calc. obs.

100 (.0277 ) W ) 130 ) ) .0279 ) .0278 .0278 W ) 001 - - .0323 - 1 101 .0601 .0601 .0601 M 181 110 (.0826 ) .0832 .0834 W 217 ) .0837 ) M 5 - .0938 - _ -a - - .0979 - - MU - 200 .1107 ) .1113 .1112 MS 289 .1120 M 5 111 (.1151 ) .1156 .1157 S 1000 .1162 ) VS - .1218 - - VW - 002 .1300 .1300 .1292 S 313 201 .1432 .1438 .1435 S 342 .1445 ) M - .1539 - - VW - 102 .1577 .1577 .1570 VVW 26 - .1628 - - VW - 174. TABTR XXVI cont:

2 average 2 hkl sin 0 sin 0 I I obs.. sin2 e calc . obs. calc. obs.

210 .1945 .1945 .1946 VVW 19 112 .2136 .2136 .2126 W 119 .2222 - - VW 211 (.2255 .2266 .2269 VVW ) 47 .2277 VVW ) 202 .2405 .2405 .2404 MW 190 300 (.2496 .2509 .2502 VVW ) 47 .2523 VVW 301 (.2814 .2826 .2825 MW ) 264 .2839 MW ) 003 .2907 < 1 103 .3194 16

VS = very strong, S = strong, MS = medium strong M = medium, MW = medium weak, W = weak, VW = very weak, VVW = extremely weak. 175.

TABLE XXVII 0 Summary of the closest interatomic distances in A

Th-Th Th-X X-X

Th7Ru3 3.44 2.85 ThRu 3.75 2.87 2.86 3.31 3.18 2.70 ThRu2

Th7Rh3 3.44 2.87 - ThRh 3.78 2.92 2.92 3.96 2.67 3.05 p-ThRh2 ThRh 4.16 2.93 2.93 3 3.26 3.10 2.98 Th2Pd Th Pd 3.93 3.03 2.87 3 5 ThPd 4.18 2.92 2.92 3 4.11 2.90 2.90 ThPd4

Th70s3 3./m 2.87 3.34 3.20 2.72 ThOs2

Th71r3 3./m 2.88 ThIr 3.79 2.94 2.93 3.32 3.18 2.70 ThIr2 /ThIr 4.29 3.07 2.64 5

Th Pt 3.47 2.90 3 ThIt 2.84 3.01 2.99 176 Auendix 2. Nominal compositions of the alloys examined and critical molting data. Th-Ru System. Annealing Presence of Annealing Presence. of % Ru Temp8rature incipient Temp rature Lncipient meltina_ -a, luelting 1 10, 15 1250 Hb 1275 Yes 20 1200 No 1300 Yes 30 1400 No 1425 All melted 32 1375 No 1400 Yea 33.3 35 37 39 1450 All melted 40 1400 No 1425 No, Yes * 43 44 1375 No 1400. Yes 50 1450 No 1475 All melted 52 1425 No 1450 Yes 53 57 60 1450 Yes. 66.6 1500 No 70 1500 No 75 83 90 * Refers to duplicate experiments. 177

The Th—Rh System.

Annealing Presence of Annealing, Presenceof % Rh Temperature: incipient Temprature dricipient oc melting 'C melting 1 15 1250 Yes 20 1225 No 30 1350 No 1375 All melted 33.3 37 1300 No 1325 Yes 45 1300 No 1325 Yes 50 1500 No 53 1475 No 1500 Yes 55 56 57 58 1425 No 59 1425 No 1450 Yes 60 1450 No 1450 Yes 62 63 1400 No 64 1425 No 1425 Yes 65 1400 No 66 66.6 1500 No 70 73 75 77 1500 No 83 85 86 89 1450 No 1475 Yes 95 1450 Yes 178

Th-Pd System

Annealing Presence of Annealing Presence of % Pd Temp8rature Incipient Temp5rature incipient melting melting

1 15. 1100 No 1125 Yes 20 1150 Yes 25 30 33.3 1150 No 1175 Yea. 35 36 38 40 1125 No 1150 Yes 43 1150 Yes 46 1100 No 50 1400 No 1425 completely molten 51 1300 No 1325 Yes 55 1300 No 1325 No 56 57 1275 No 60 1200 No 1225 Yes 61 1200 No 1225 completely molten 63 1300 No 1400 Yes 66.6 1350 No 68 1375 No 1400 Yes 70 75 1500 No 77 78 79 1350 No 80 81 82 83 1250 No 1300 Yes 84 85 86 1150 No 1175 Yes 87 1125 No 1150 Yes 88 1100 No 1125 Yes, No 89 1100 No 1125 Yes 90 1150 No 1175 Yea 95 1300 No 1350 Yes 98 179

The Th-Os System

Annealing Presence of Annealing Presence of %Os. Temp rature incipient Tem: erature incipient 'C melting oC melting

1 10 15 1275 No 20. 1250 No 1300 Yes: 30 1500 No 33.3 35 1475 No 1500 Completely molten 37 1450 No 39 40 45 47 1500 No 50 661.6 75 83 1500 No 95 1500 No 180 The Th-Ir System.

Annealing Presence of Annealing Presence of %Ir Temperature incipient Temperature incipient oG melting. melting,

1 10 1350 Yes 15 1325 No 20 1300 No 30 1500 No 33 c, 3 35 1450 No 1500 Yes 36 1475 Yes 37 1475 Yes 40 45 50 53 1500 55 57 60 63 66 68 70 1500 No 72 75 77 150o No 80 8l 83 86 1500 No 88 90 95 1500 Into 181.

Th—Pt System.

Annealing; Presence of. Annealing Presence of % Pt Temperature incipient Temperature incipient oa melting C melting 1 10 1250 Yes 15 1225 No 17 20 1200 No 1275 Yes 25 3.0 33.3 1350 No 1375 Yes 40 1350 No 50 53 55 1500 No 56 5? 60 1500 No 6.3 65 66 69 70 1500 No 75 77 1500 Nb 80 81 1500 No 83 85 1325 No 1350 Yes 86 1300 No 90 94 1325 No 1350 completely molten 95 1350 completely molten

>1500 1700 N it, % / 1 % / I 1 % / t 1 1 / 11 / 1600- ' / 1 1 / 1 L j/ I 1 >1500- - 1 1500- 1462t12 c `‘I 1425+ 12 / 1438±12 1412 4:12 /11 / 1400- 13 88 - -

0 g..) 1300 - 1 II 4J 1262 t 12

E 1200 ty

1100 N 3 CC CC rf .a 1- 1- 1000 20 40 60 80 100 Atomic ale ruthenium.

Figure 1. The thorium -ruthenium system. >1500 1700

/ 1600 1t / I / t / I 1

1500 t I /407 I t +12 I 14501' 12 I I 1 1425 1400- t ± 12 1 1362±121 - - - - -.\ I / \1 V 1 0 t // 1300- 13121 12 V 1 / t- 1 i = 44 li Of 1237 ± 12 L 1200 E

1100- cr

1000- 1 0 20 40 60 Atomic 0/0 rhodium. Figure 2. The thorium-rhodium system.

>1500 1700. ill % Ii ' % / i t I I I I i 1600 I / I 1 I 1 I 1 1 1 I 1500- i i t 1 I i 1 1412± 12 1 r \ 1400- 1 I 1 I ‘ 1 I N 1 1337 ‘ I + 12 --- -A t 6 1 / 0 / 132 \ I 1 ±121 1300- i 1 1.- 4-0 i q M I 1 1212 41) 1200L / 412 •itt E 1 1162.432 t 13 t1 1 I/ 1137±12 I- 1125 12 1100 - 1112 12 •st in #.13 .0 0) st. a_ EL 0) ce) EL .0 .c ,..c 1- i- 1- t, 1000, 1 L.) 20 40 60 100 Atomic O/ palladium.

Figure 3. The thorium-palladium system.

>1500 >1500 1700,

1600- 1 I I 1 >1500 I I 11 1 >1500 I I 1 1500- 1 1 I 1487 1 I ±12

1400- 1 1 1 ---.1 I Lj 1 I I f 0 a) 1300 V 1287 ± 12 4-0

1200- E

1100- ol Nu) x u) 01, 0 0 .c t .c I- I- I- . I 1000. 20 40 60 80 Atomic 0/0 osmium.

Figure 4. The thorium -osmium system. >1500 1700

1600-

1500

114621: 12 14621:12 1400

QU 1337 ± 12 2.! 1300 m L a E 1200

1100- LT t, L.. LN .c .c .0 I-- I- 1- 10000 20 40 60 80 100 Atomic 0/0 iridium.

Figure 5. The thorium-iridium system. 1700\

1600 -

1500

I / 1400 1

MEI ••• 13 6 2 ± 12 0 I / / 1337 ± 12 w 1300 1 / L I / +.0 co 11 L cu 12 37 ± 12 X1200

1100- ro I' Ul 4•J .4..' 44 V in a. 4-1 CL EL t-• cr. pi et .G .0 ..0 C I— I— I— I— I- 1000 t I 20 40 60 80 100 Atomic °A. platinum..

Figure 6. The thorium-platinum system. ThSe 1600 •

01-5 tu 1400

-+Th7lri • MN. -*Th70s3 Th2AI Th Ft?Th 7Ru 3 . -0.Th Bei3 1200— Th7Rh3 Thjln .Y • Th2Pd ThiNi 3 1000— Th.,Cu

J i J , 1 i I ...... JL-____i 1000 1200 1400 1600 '1800 Melting point of the compound. °C

Figure.7.

Th Ni -.Thae13 28 3

.4, 26 0 Th2Cu • fo c 24 0 7-; Ti Th2Pd a 22 • E (.)o Th2A1 20 • • Th7Rh3 Th2ln 7 18 • Th7Pt3 LL.1 Th7Ru3 16 • Th71r3 ThSe -40 • T hP S3 14 1000 1200 1400 1600 1800 Melting point of compound ° C

Figure. 8.

Se 1600 X

Ir x Ru x Pt x Os Al X x x u 1200 Rh Be x Pd In x

1000 x Ni Cu x

3.4 3.6 3.8 4.0 4.2 Mean electron concentration.

Figure.9. 3.4

TI 3.2 •In o 3.31

3-0 O. Au Al Ai • Ga • 3-20 ter 317 _ n 3.22 2-8 1.....1 Pd me • ia Zn 3.13 • d

ic 319 2-6 Cu •

Atom 3.11

24 I I I i I_ 1.3 1.5 1.7 1.9 2-1 2.3 Electronegativity of X

Figure10. Interatomic distances between unlike atoms

in Th2X compounds.