History of Niobium

The discovery of niobium is intimately connected with that of tantalum, firstly because these metals are consistently associated together in their natural ores, and secondly, because their separation from one another has proved an extremely difficult matter. Indeed, the chemistry of these elements is so closely parallel that considerable time elapsed before their separate identities were definitely established.

In 1801 Hatchett, an English chemist, while working on some chromium minerals in the British Museum, examined a black mineral which had been found in the Connecticut valley and had been sent to the then president of the Royal Society. Hatchett reported the existence in the mineral of a new element, for which he suggested the name columbium, because the ore had come from America. The mineral itself became known as columbite. In the following year Ekeberg found what he thought was another new element in two other minerals, what is now known as tantalite, from Kimito (Finland), and yttrotantalite, from Ytterby (Sweden). This new element was named tantalum, partly because mythological names were then common for new elements, and partly because of the tantalising difficulty that was experienced in dissolving the oxide of the new metal in excess of acids. Wollaston3 subsequently compared the properties of the two new elements by re-examining their ores, and endeavoured to show that they were identical. It is now known that what was then looked upon as the pure oxide must have been tantalum pentoxide mixed with small proportions of niobium pentoxide, and the new " element " was therefore impure tantalum.

In 1839 Wohler investigated some peculiar properties of an acid oxide (now known to be a mixture of niobic acid and titanic acid) present in the mineral pyrochlore and in some Bavarian tantalites. Rose followed up the observation that many columbites and tantalites, as well as the acids obtained from them, displayed widely varying densities, and, after close investigation into their composition, he announced in 1844 that the columbites from Bodenmais in Bavaria contained, in addition to tantalic acid, the acid oxide of a metal which was not present in columbites from Sweden and Finland. The new metal was named niobium after Niobe, a daughter of Tantalus. In 1846 Rose thought that he had discovered the acid oxide of still another metal, to which he gave the name pelopium, but later he decided that this new acid was merely hyponiobic acid. In 1856-57 Hermann showed that both tantalum and niobium occur in the various natural tantalites and columbites.

From a consideration of the composition of the halides, and because of the supposed isomorphism of tantalic acid and stannic acid, Rose gave the formulae NbO₂ and TaO₂ to anhydrous niobic acid and tantalic acid respectively, but in so doing he repeated an error that had previously been made by Berzelius with regard to vanadium compounds, and overlooked the presence in the halides of an oxygen atom. Blomstrand and Marignac subsequently analysed a large number of chlorides of different origin, and investigated the constitution of the fluorides and double fluorides of niobium. A "hyponiobic fluoride" was found to contain three atoms of fluorine for one atom of oxygen, and its double fluorides with the fluorides of other metals were shown to be isomorphous with similar double fluorides given by titanium tetrafluoride, TiF₄, tin tetrafluoride, SnF₄, and tungsten oxydifluoride, WO₂F₂. Furthermore, a comparison of the oxyfluorides of tungsten with the fluorides of titanium, tin, silicon and zinc had previously shown that an atom of fluorine in these compounds could be replaced by an atom of oxygen without disturbing the isomorphism. The formula for " hyponiobic fluoride " thereupon became NbOF₃, and in consequence, niobie anhydride became Nb₂O₅. The formula for tantalic anhydride was established as Ta₂O₅ at the same time, because tantalic acid occurs in isomorphous mixtures with niobic acid in several minerals, from which isomorphous double fluorides, such as potassium tantalum fluoride, K₂TaF₇, and potassium niobium fluoride, K₂NbF₇, were also obtained. These conclusions were confirmed by determinations of the vapour densities of niobium pentachloride, NbCl₅, tantalum pentachloride, TaCl₅, and niobium oxytrichloride, NbOCl₃, by Deville and Troost in 1865.

Marignac also showed that previous methods for the separation of niobium and tantalum were far from perfect, and for the first time he succeeded in preparing pure niobium and tantalum compounds. His methods are still in use to some extent, and his analyses provided the first reliable values for the atomic weights of these elements. It should be stated, too, that Rose's earlier researches, which extended over a period of nearly twenty years, have provided a valuable source of information for the chemistry of niobium and tantalum. His calculations and formulae were revised by Rammelsberg in the light of subsequent discoveries.

During the years 1860 to 1871 the presence of various other elements in columbites and tantalites was reported. Hermann defended the original formulse of Rose for niobium and tantalum compounds, and announced the existence of another element, which he styled ilmenium, in yttro-ilmenite, samarskite, and other niobium-bearing minerals, but ilmenium was shown by Marignac to be a mixture of niobium and titanium. Hermann also claimed the discovery of still another metal, which he called neptunium, but Blomstrand and Larsson independently showed it to be identical with niobium, and Smith was unable to confirm Hermann's preparation. Kobell came to the conclusion that niobites from Bodenmais in Bavaria contained the acid oxide of a metal which he named dianium, and which was not present in niobites from North America and Greenland; but the identity of dianium and niobium was also soon established. Kriiss and Nilson, on spectroscopic evidence, assumed the existence of still another element in fergusonite, but the assumption became untenable. A re-investigation of American niobites and tantalites carried out in 1905 failed to bring to light any new elements other than niobium and tantalum; this result was confirmed by spectroscopic examination of several samples of niobium pentoxide which had been prepared from various minerals.

Niobium and tantalum suddenly received considerable attention about the year 1905 as possible materials for the filaments of incandescent electric lamps in place of the carbon filament then in use. The metals were then prepared in the pure state for the first time by Dr Werner von Bolton, and their properties were examined. Niobium was found to be unsuitable for the purpose in view, but tantalum proved to be satisfactory. Tantalum lamps were manufactured in large quantities between the years 1905 and 1911, when the metal was displaced by the electrically more efficient tungsten.