VI-The structure of some ternary alloys of copper, By A. J. BRADLEY, M.Sc., Ph.D., and C. H. GREGORY, M.Sc. (Communicated by Professor W. L. Bragg). SOLID SOLUTIONS IN INTERMETALLIC COMPOUNDS. Our present investigations have been undertaken in the hope of throwing light on the laws governing the formation. of certain so-called intermetallic compounds. It is well known (cf. Desch. Intermetallic Compounds; Longmans, Green & Co., 1921) that many intermetallic compounds have formulæ which do not fit in with the ordinary ideas about valency. This is especially the case with compounds which contain only highly electropositive metals. Recent investigations using X-ray methods have shown the existence of a more serious difference between many of these compounds and the normal type of chemical compound. The law of definite proportions which has been established for over a century as the direct consequence of experimental evidence states that "a compound always contains the same elements combined together in the same proportions by weight." The universality of this law has led to its being accepted as a criterion for the existence or non-existence of a compound. X-ray investigations have now shown conclusively that there exists a class of metallic alloys which possess many of the attributes of a compound, but which do not obey the law of definite proportions. This kind of alloy is usually known as a solid solution in an intermetallic compound, but this designation gives a wrong mental impression of the nature of these bodies, as it presupposes the existence of a definite compound with characteristic chemical and physical properties as the basis of the solid solutions phase. From the results of recent work it appears probable that phases of this type are simply space lattices of a special type. They are in fact mixtures of their respective ingredients in certain proportions which may vary between wide limits. In many cases these limits are fixed by laws which are quite unconnected with the laws of definite proportions or with the laws of valency as will be seen from the results quoted in the present paper. Tammann1 has suggested that in solid solutions the atoms will tend to sort themselves out into separate space lattices. This effect should occur on annealing any solid solution, but X-ray investigation has shown that in practice, this sorting out process only occurs very rarely. For example, Johansson and Linde2 have found that alloys of copper with gold, palladium or platinum in certain proportions, show this effect. For example, an alloy containing 25 atomic per cent. gold and 75 atomic per cent. copper, before annealing has the facecentred cubic lattice typical of the parent elements, with a haphazard distribution of the constituents. After annealing, the gold and copper atoms are sorted out into definite positions. The ultimate effect is to yield a theoretically perfect distribution of atoms, the copper atoms all lying on one lattice and the gold atoms on the other lattice, so that the structure is exactly like that of a normal compound of the ionic type. The annealed alloy has therefore the structure of a compound with the formula AuCu,. An imperfectly annealed alloy of this composition on the contrary, is evidently a solid solution with. a tendency to regular arrangement. In general, phases which are commonly termed solid solutions in intermetallic compounds are analogous to the imperfectly arranged gold-copper alloys. The tendency to regular distribution is due to the operation of several independent factors. Of these the most important are the existence of geometrical inequivalence between the atomic positions, and of a fairly high degree of difference between the properties of the alloying elements. Annealing is only necessary when these two factors are inoperative. If the proportions of the ingredients correspond exactly 1. Lehrbuch der Metallographie, 2nd Edition, Leipzig, 1921, 2. Ann. der Phys. 78, 429 (1925), 82, 449, (1927). p. 329. to the geometrical requirements, the distribution may be complete, giving an ideal intermetallic compound, using this term as it is defined by Westgren and Phragmèn.1 With any other concentration the distribution is necessarily incomplete. The alloy is then in an intermediate state, lying between an ideal solid solution on the one hand and an ideal intermetallic compound on the other hand. The terms ideal solid solution and ideal intermetallic compound have been defined by Westgren and Phragmèn in the following way: In an ideal solid solution all the atoms are structurally equivalent. In an ideal intermetallic compound structurally equivalent atoms the chemically identical. The intermediate type of alloy, which is usually known as a solid solution in an intermetallic compound, would perhaps be better described as a complex solid solution, defining the term in the following way : A complex solid solution consists of a single phase, the relative amounts of the components being variable, with the atoms not all structurally equivalent. The The study of a specific alloy system shows the application of this definition. The system Ag-Zn has been investigated by Westgren and Phragmèn and others. There are five phases known as a, ß, y, ɛ and ŋ. phase consists of silver with any proportion of zinc up to 27 per cent. The structure is identical with that of pure silver, being facecentred cubic. Silver and zinc atoms are distributed quite haphazard throughout the crystal, so that the probability that an individual atom will be zinc rather than silver is exactly proportional to the number of zinc atoms present. Thus, taking the structure as a whole, it behaves as if all the atoms were alike, and is therefore an ideal solid solution. The phase is body-centred cubic but the atoms are not all equivalent, the atoms at the centres of cubes and the atoms at the corners of cubes being unequal in their capacity for scattering X-rays, as is shown by the X-ray diagrams. Since 1. Phil. Mag., 50, p. 31 (1925). |