Disorder and Order in Solid Solutions of Oxygen in alpha-Titanium.

Abstract

The range of solid solubility of oxygen in a-titanium has been studied by means of X-ray powder and single-crystal techniques. Three phases have been observed, viz. a random solution (TiOo+�) and two phases based on an anti-Cd(OH) 2 structure type. One of the latter (Ti 2 01-y) has a broad range of homogeneity, the oxygen vacan­ cies indicated in the formula being randomly distributed. The other, (Ti20i-'/,), has one third of the oxygen positions of Ti20 vacant in an ordered manner. The occurrence of the three phases was found to depend on both the oxygen content and the annealing temperature. T he hexagonal close-packed a-modifications of titanium, zirconium and hafnium are characterized by a remarkable ability to dissolve small nonmetal atoms to a large extent. This is especially so with oxygen which can enter into solid solution up to the compositions Ti0 0_50 h', Zr0 0. 40 5 and Hf0 0•26 6 in the three metals, respectively. The solubility mechanism has been thought to imply that the oxygen atoms occupy the major interstices of the metal structure, i.e. the octahedral holes, in a perfectly random way 7 , 8 • Studies recently carried out at this Institute have, however, demonstrated the existence of ordering at higher oxygen contents in the a-zirconium solid solutions 5 while a random arrangement is present in a-hafnium solid solu­ tions 6 • According to previous investigators, the incorporation of oxygen atoms in the a-titanium and a-zirconium structures causes a considerable expansion of the c axes while the a axes are less affected by the oxygen content. It was, however, found by the present research group that the length of the a axes actually passes through a maximum value occurring at the composition Ti0 0 • 33 1 and Zr0 0•25 5 , respectively. At these contents of oxygen, there is also observed a marked acceleration of the rate of increase of the c axes. The present investigation of the a-titanium-oxygen solid solutions has revealed that several preparations may give extra X-ray reflexions which are not compatible with the idea of random solutions of oxygen. The various samples employed in these studies were prepared in the way described in a Acta Chem. Scand. 16 (1962) No. 5