Page images

A. Lacroix, G. Spezia, A. Schenk, F. Kollbeck, J. S. Diller, O. Mügge, B. Doss, J. B. Scrivenor, H. Rösler, C. Schmidt, K. von Chrustschoff, etc. H. Thürach mentioned the occurrence of brookite and anatase in a great variety of rocks. Brookite alters into rutile; and rutile into ilmenite, and sphene. The occurrence of brookite has been discussed by A. Müller, J. R. Blum, A. Lacroix, etc.

The preparation of titanium dioxide.-Amorphous titanium dioxide was made by F. Wöhler,12 and H. Rose by adding aq. ammonia to a soln. of titanium tetrachloride, washing and calcining the precipitate; it is also made by the processes indicated in connection with the extraction of titanium-vide supra-namely, by fusing rutile or titanic iron ore with potassium carbonate, decomposing the mass with hydrofluoric acid; separating and crystallizing the resulting potassium fluotitanate, and decomposing its hot soln. with aq. ammonia; or by heating titanic iron ore in a stream of chlorine and hydrochloric acid-the iron is volatilized as chloride, and the titanium dioxide remains: 2FeTiO3+4HC1+Cl2=2FeCl3 +2TiO2+2H2O. H. Geisow described the extraction of titanium dioxide from rutile, etc.; and C. Winkler, and G. Cartaret and M. Devaux, the purification of commercial titanium dioxide. J. J. Ebelmen showed that if the amorphous titanium dioxide be heated with boric acid, or better, with microcosmic salt, in a pottery oven, golden yellow, acicular crystals of rutile were formed. According to G. Wunder, the crystals are sodium phosphotitanate, and not anatase as G. Rose assumed, or titanyl phosphate, TiO(PO3)2, as A. Knop supposed; on the other hand, R. Brauns believed them to be titanium sesquioxide, and L. Ouvrard, rutile with titanyl phosphate and alkali titanophosphate. G. Rose, and A. Knop obtained fine crystals of rutile by heating for a long time amorphous titanium dioxide mixed with borax, or with borax and microcosmic salt. B. Doss used borax. J. J. Ebelmen also obtained rutile by heating the amorphous dioxide with potassium carbonate. H. St. C. Deville and H. Caron prepared rutile by heating to redness a mixture of amorphous titanium dioxide, silica, and stannic oxide. L. Bourgeois used barium chloride as a flux; F. A. Genth, potassium hydrosulphate; and P. Hautefeuille, sodium silicate, tungstate, or vanadate; H. Cormimboeuf melted titanium dioxide with sodium carbonate, sodium tungstate, and as much tungstic oxide to make the composition of the fused tungstate that of the normal tungstate-crystals of rutile are formed; with less tungstic acid, crystals of the alkali titanate are formed. H. Traube also used sodium tungstate as agent minéralisateur, and he was able to add to the rutile appreciable quantities of iron, manganese, and chromium which are present as impurities in the natural mineral. T. Scheerer observed crystals of rutile in the cracks of the brickwork of a blast-furnace; and K. Endell, in pottery glazes containing 5 to 10 per cent. of rutile. H. E. Merwin and J. C. Hostetter obtained crystals of rutile by the action of chlorine on titaniferous clays at 1000°-1100°. F. Wöhler heated titanium nitride in air, and obtained crystals of what were thought to be rutile. L. Michel obtained a mixture of rutile and pyrrhotite by heating a mixture of ilmenite and pyrite. H. de Sénarmont heated amorphous titanium dioxide in the presence of carbon dioxide in a sealed tube at 200°. A. Daubrée decomposed titanium tetrachloride by steam at a red heat and obtained crystals of the dioxide; and H. St. C. Deville passed hydrogen fluoride or chloride over red-hot titanium dioxide and obtained crystals of rutile; if a reducing atm. be present, blue crystals of TiO(TiO2)2 were formed. Similar results were obtained by P. Hautefeuille and A. Perrey. The three forms of titanium dioxide were studied by P. Hautefeuille. A mixture of potassium titanate and potassium chloride heated to redness in a stream of dry air mixed with hydrogen chloride furnished crystals of rutile. Similar crystals were obtained by heating to redness a mixture of titanium dioxide and potassium fluoride or calcium fluoride, or a mixture of titanium dioxide and potassium fluosilicate in hydrogen chloride. In these experiments, if the temp. exceeded 1040°, rutile was formed; between 800° and 1040°, brookite; and below 800°, anatase. O. Lehmann discussed the transformation of one form into another. C. Friedel and J. Guérin heated to

redness titanium chloride and ferrous oxide, and obtained ferrous chloride and crystals of titanium dioxide; and by passing chlorine over titanium-iron at a red heat, ferric chloride is volatilized and crystals of rutile are formed. They proposed the reaction as a means of separating titanium from iron.

F. Wöhler heated titanium carbonitride in a current of steam and obtained crystals of anatase; and H. Rose obtained anatase by heating amorphous titanium dioxide for a short time by means of a spirit-lamp. G. Rose found some crystals of anatase in the product obtained from a soln. of titanium dioxide in molten microcosmic salt. The process was examined by A. Knop, G. Wunder, B. Doss, and L. Ouvrard. F. A. Genth found that some anatase accompanied the rutile when titanium dioxide was fused with potassium hydrosulphate. B. Doss' attempts to make anatase with fused borax as agent minéralisateur were not successful, rutile was always formed. P. Hautefeuille obtained anatase by the decomposition of titanium tetrafluoride by aq. vapour, at or near 860°; and also by heating titanium trifluoride in a current of air. P. Hautefeuille and A. Perrey found that hydrogen chloride has no action on titanium dioxide at a bright red heat, but under a press. of 3 atm., anatase is formed at a dull red heat; at the same temp. and under ordinary atm. press., hydrogen chloride has a mineralizing effect on the carbonate, oxalate, or sulphate, but not on the oxide.

A. Daubrée made crystals of brookite by passing a mixture of the vapour of titanium tetrachloride, steam, and carbon dioxide through a red-hot tube; and also when the vapour of titanium tetrachloride is decomposed by heated lime. P. Hautefeuille found that brookite is formed when potassium fluotitanate is heated in steam, and when hydrogen fluoride acts on titanium tetrachloride at a temp. not exceeding 1040°. A mixture of titanium dioxide, calcium fluoride, and potassium heated in a stream of hydrogen chloride, silicon tetrafluoride, and moist hydrogen, also furnished crystals of brookite, and a similar result was obtained when a mixture of titanium dioxide, silica, and potassium fluosilicate was heated in a current of hydrogen chloride alone. B. Doss obtained rutile, not brookite, from soln. of titanium dioxide in fused borax.

The physical properties of titanium dioxide. Purified titanium dioxide is colourless; the colour of the mineral forms is due to the presence of impurities, is usually reddish-brown passing into red, and sometimes yellowish, bluish, violet, black, and rarely, grass-green. In transmitted light the colour is yellow, various shades of red, and violet. Anatase may be various shades of yellow and brown; and sometimes indigo-blue or black. In transmitted light, the same colours may appear, and they are sometimes distributed zonally or irregularly. Brookite may be yellowish- or reddish-brown or iron-black. In transmitted light the colour may be yellowish, reddish, brownish, or colourless, and rarely blue or bluish-green J. W. Retgers 13 said the colour of rutile is largely determined by the titanium dioxide and not by the ferric oxide. L. Wöhler and K. von KraatzKoschlau obtained colours ranging from bluish-black to greenish-black with titanium sesquioxide as the tinctorial agent. Red rutiles were obtained only with the ferric oxide; vanadium oxide had no effect on the colour. H. Traube melted mixtures of rutile with ferric oxide and obtained a dark brown product with 1.98 per cent., and a black mass with 5.4 per cent. of ferric oxide; manganese oxide gave a yellow tint, and the colour was bluish-black when 3 per cent. of this oxide was present. With chromic oxide, a green colour was obtained. R. F. Wagner said that the colour of titanium dioxide is determined by the treatment it has received. When moist orthotitanic acid is heated, there is a remarkable play of colours. The oxide becomes almost white when cold and citron-yellow when hot, and by continued heating the colour becomes more and more brownish.

O. Hahn 14 studied the surface area of the particles of different forms of precipitated titanic oxide in terms of adsorbed radioactive matter. The habit of the euhedral crystals of rutile is commonly prismatic with furrows or striations on the surface parallel to the c-axis. The crystals are often acicular, very slender,



or hair-like. Transparent quartz is sometimes penetrated thickly with acicular or capillary crystals as illustrated by Fig. 6. This furnishes the so-called sagenite and crispite alluded to above; and the veneris crinis of Pliny's Historia Naturalis

FIG. 6.-Rutile Needles in Quartz.

and the so-called flêches d'amour or Venus' hair stone are varieties of quartz penetrated with acicular rutile. Rutile occurs compact and massive, and in granules. Anatase is usually octahedral, in habit acute or obtuse, and also tabular. It is rarely prismatic. The habit of brookite crystals is varied. The euhedral crystals are often tabular parallel to the (100)-face; with the (100)-face and the prismatic faces striated vertically. Sometimes the habit is prismatic with the (110)face dominant, resembling rutile. According to R. J. Haüy, crystals of rutile belong to the tetragonal system; and more exact measurements were made by A. Breithaupt, G. A. Kenngott, and W. H. Miller. The last-named gave for the axial ratio a: c=1: 0-644154; H. Baumhauer, 1:0.6439; and N. von Kokscharoff, 1:0-64418. The optical anomalies of rutile led E. Mallard to assume that it does not belong to the tetragonal system; and F. Wallerant regarded rutile as monoclinic. G. Friedel, and W. J. Sollas considered the space-lattice to be tetragonal. Observations on the crystals of rutile were made by H. Baumhauer, R. L. Parker, G. vom Rath, F. Hessenberg, C. O. Trechmann, F. Pisani, O. Lincio, H. Tertsch, A. von Lasaulx, W. C. Brögger, P. von Jeremejeff, G. Tschermak, F. Rinne, D. Brewster, O. Mügge, M. Bauer, L. Cahn, G. Rose, H. S. Washington and W. E. Hidden, etc. The crystals of anatase were examined by R. J. Hauy who placed them in the tetragonal system and the cognomen, as previously indicated, implied that the c-axis is longer than is the case with rutile. W. H. Miller gave for the axial ratio a: c=1:1-7771. Observations were made by C. Klein, A. des Cloizeaux, O. Luedecke, C. Vrba, A. Frenzel, A. Stelzner, R. L. Parker, O. Pohl, F. Wiser, G. vom Rath, A. Brezina, G. Boeris, A. Lévy, H. Buttgenbach, J. Beckenkamp, G. Cesaro, R. W. Haare, O. C. Farrington and E. W. Tillotson, H. P. Whitlock, W. Prinz, J. Schetelig, S. L. Penfield, L. Colomba, A. Johnsen, H. Baumhauer, F. Millosevich, G. Seligmann, A. Sella, A. Streng, E. Bertrand, F. Hessenberg, C. Busz, A. E. Robinson, etc. The crystals of brookite, said F. S. Beudant, were confondue pendant long-temps avec le rutile. A. Lévy found that they belonged to the rhombic system. N. von Kokscharoff gave for the axial ratios a: b: c=0.84158:1:0-94439. The crystals were measured by K. Romanowsky, W. H. Miller, and A. des Cloizeaux. A. Schrauf tried to show that the crystals of brookite are really monoclinic, but both N. von Kokscharoff and P. Groth showed that this is not a correct interpretation. Further observations were made by C. Vrba, G. Lechner, A. Wichmann, F. Wiser, C. Palache, A. Fornaro, L. Brugnatelli, G. vom Rath, C. Busz, G. Rose, P. von Jeremejeff, G. F. Kunz and S. L. Penfield, A. Breithaupt, E. S. Dana, etc. In addition to the three forms, rutile, anatase, and brookite, R. B. Riggs and J. S. Diller reported a fourth rhombohedral form occurring in thin iron-black scales as inclusions in the tourmaline of Hamberg, New Jersey, and De Kalb, New York. J. D. Dana, however, regarded it as a variety of ilmenite.

The twinning plane and composition plane (101) of rutile are often geniculated, and there are contact twins of very varied habit. There is also polysynthetic twinning in thin lamella parallel to (101), and the lamellæ are of various lengths,


and distributed irregularly in the crystals. The cleavage of rutile parallel to (110) and (100) is distinct; and that parallel to (111) is in traces. This subject was studied by A. Breithaupt, G. A. Kenngott, W. H. Miller, etc. O. Mügge 15 reports a parting due to twinning parallel to (902). The cleavage of anatase parallel to (001) and (111) is perfect; and those of brookite parallel to (110) and (001) are indistinct and seldom observed in microscopic crystals. The corrosion figures of rutile with molten potassium fluoride, or hydrofluoride, were studied by H. Traube. Naturally etched crystals of anatase, and crystals etched with potashlye were studied by H. Baumhauer; and the gliding planes of rutile, by A. Grühn and A. Johnsen. Observations on the optic axial angle of brookite were made by A. Beer, 16 W. H. Miller, J. Grailich and V. von Lang, etc. J. Grailich gave · 2E-65° for the red ray; and A. Schrauf gave:




[ocr errors]


. 58° 22′ 55° 21'

661 592 589 555 534 532 486 50° 45' 28° 56' 30° 1610 33° 48' 38° 7' c. 70°

A. des Cloizeaux measured the effect of temp., and found that by raising the temp. to redness, 2E changed from 42° to 47°; and U. Panichi changed 2E from 54° to 26° by cooling to


Anatase is isomeric but not isomorphous with rutile. The X-radiogram has been studied by L. Vegard, 17 H. Haga and F. M. Jäger, W. P. Davey, J. Beckenkamp, H. Tertsch, G. Greenwood, A. Johnsen, R. L. Parker, M. Born and O. F. Bollnow, and C. M. Williams. According to L. Vegard, in the space-lattice of zircon, ZrSiO4, the oxygen atoms are arranged in pairs about the zirconium and silicon atoms, and he considers that the constitutional mol. formula of zircon is accordingly ZrO2.SiO2, not ZrSiO4. The space-lattice of anatase can be derived from that of zircon by removing the zirconium atoms and their associated oxygen atoms, and substituting titanium atoms for the silicon atoms. The oxygen atoms are arranged in line with the titanium atoms with the lines parallel to the tetragonal axis. Hence, L. Vegard writes the mol. formula TiO2. The estimated absolute dimensions of the space-lattices of rutile and anatase are given in Table II. C. M. Williams's data do not agree with those of L. Vegard, and he TABLE II.-DIMENSIONS OF THE SPACE-LATTICES OF RUTILE AND ANATASE.

[merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small]

assigns to the crystal unit a different structure. According to M. L. Huggins, in rutile and anatase, each titanium atom is surrounded by four equidistant oxygen atoms, and each oxygen atom by two equidistant titanium atoms, all at tetrahedron corners. F. Rinne represents the arrangement in the tetragonal unit by the drawing, Fig. 7. The observed data are not yet sufficient to establish the structures of anatase and rutile.

FIG. 7.-The Tetragonal Unit of Rutile-F. Rinne.

G. Rose 18 assumed that rutile, zircon, and cassiterite are isomorphous; and P. Groth wrote the formula TiTiO, and SnSnO4, in order to emphasize their relationship to ZrSiO4. He also considered brookite to be related to tridymite, and anatase to cristobalite. J. W. Retgers did not agree with the assumption that these minerals are isomorphous, and maintained that there is a chemical contrast between zirconium and silicon which is not characteristic of isomorphism; titanium and silicon dioxides, said he, nicht die geringste Neigung zu inniger Mischung zeigen

this is illustrated by the rutile needles found in colourless crystals of quartz. Further, tin dioxide does not appear to be miscible with titanium or zirconium dioxide. Against J. W. Retgers' view, H. Traube found that the corrosion figures of rutile, zircon, and cassiterite show the holohedral tetragonal symmetry. Some crystallographic data for the minerals rutile, TiTiO4; zircon, ZrSiO4; cassiterite, SnSnO4 ; thorite, ThSiO4, polianite, MnMnO4; and plattnerite, PbPbO4, by S. Stevanovic, are shown in Table III. The mol. vol. axis ratios, and the topic axes increase with the TABLE III.-CRYSTALLOGRAPHIC PROPERTIES OF THE ISOMORPHOUS DIOXIDES.

[blocks in formation]

mol. wt. except in the case of polianite, which probably does not belong to the series, as is also the case with silica. H. Buttgenbach, V. Goldschmidt, A. Schrauf, G. Wunder, R. Ruer, G. Linck, W. T. Schaller, and M. Ladrey emphasized the crystallographic relationshp between the dioxides of tin, titanium, and zirconium. O. Lehmann discussed the transformation of one modification of titanium dioxide into the other.

M. H. Klaproth 19 gave 4.180 for the specific gravity of titanium dioxide and C. J. B. Karsten gave 3.9311. H. Rose found that when heated to 600°, the sp. gr. of the amorphous precipitate rose from 3.89 to 3.95; after exposure to a stronger source of heat for a longer time at 800°, 4·13; and after vigorous calcination at 1000°1200°, 4-255. P. Hautefeuille found that after fusion and pulverization, the sp. gr. was 4.1. H. Fizeau said that artificially prepared rutile has the same sp. gr. as the mineral. For rutile, C. M. Kersten gave 4-242; A. Virlet, 4-325-4-246, and after fusion, 4-241; M. Weibull and A. Upmark, 4·2; R. Böttger, 4-249; T. Scheerer, 4-244-4.245; A. Breithaupt, 4-250-4-291; H. Kopp, 4-420 at 0°; H. Müller, 4-56; A. von Lasaulx, 4·173-4-278. H. Geisow gave 4.21 for the sp. gr. of the strongly calcined oxide. For artificial rutile, J. J. Ebelmen found 4-260-4-283; and P. Hautefeuille, 4.3. A. des Cloizeaux and A. Damour found the sp. gr. of rutile rose from 4.273 to 4·365 when heated in a current of hydrogen. Rutile containing appreciable amounts of tin or iron oxides has a higher sp. gr. For instance, W. B. Smith found a sp. gr. 4.288 for a sample with 3.77 per cent. of ferrous oxide; F. A. Genth, 4-249 for a sample with 6.68 per cent. of ferric oxide; and W. P. Headden, 5-294 for samples with 7.92-8.10 per cent. ferrous oxide. L. N. Vauquelin gave for the sp. gr. of anatase, 3.857; N. von Kokscharoff, 3-815; E. Hussak, 3.794; F. Mohs, 3-826; A. Breithaupt, 3.750; C. Klein, 3·83–3·97 ; A. des Cloizeaux, 3.87; F. von Kobell, 3-82; H. Rose, 3-890-3-912; and A. Damour, 4.06. For artificial anatase, P. Hautefeuille gave 3.7-3.9. The sp. gr. of brookite by H. Rose ranged from 4.128 to 4:166, and on calcination, the sp. gr. became the same as that of rutile; A. Breithaupt gave 3-952; C. F. Rammelsberg, 3-892-3-949; A. Damour, 4-030-4-083; J. D. Whitney, 4-085; W. von Beck, 4-200; M. Frodmann, 4-220; K. Romanowsky, 4·1-4-2; R. Hermann, 3-83; and N. von Kokscharoff, 4-1389-4-1410. P. Hautefeuille gave 4.1 for artificially prepared brookite.

J. D. Dana 20 tried to show that there is a simple relationship between the molecular volume and the crystallographic constants of minerals; and he concluded that the relationship depended on mol. vol., and had nothing to do with chemical composition. G. T. Prior applied the theorem to estimate possible

« PreviousContinue »