The enriching of the nickel crude matte

In former times, in some localities, the crude mattes had so low a proportion of copper and nickel that the matte was subjected to a preliminary concentration before being refined. This was done by roasting the crude matte in heaps, stalls, shaft furnaces, reverberatory furnaces, or muffles, and again smelting it with a suitable flux - quartz, glass, or slag - in shaft or reverberatory furnaces. This treatment removes a relatively large proportion of iron from the matte. The subject was discussed by E. F. Kern and H. W. Walter, and H. W. Walter. C. Schnabel thus described the chemical changes involved in the smelting of the roasted matte:

Ferric oxide is reduced to ferrous oxide by the undecomposed sulphide of iron present; this passes into the slag, and sulphur dioxide is formed; a further quantity of sulphide of iron is changed to ferrous oxide by the nickel and copper oxides, and also slagged, the oxides of these metals meanwhile becoming sulphides; further copper oxide and undecomposed copper sulphide give metallic copper and sulphur dioxide. This copper is reconverted into sulphide at the expense of some iron sulphide. The iron thus separated is partly taken up by the matte, partly acts upon a corresponding amount of ferric oxide, giving more ferrous oxide to be slagged. There is no action between nickel oxide and sulphide. The various metallic sulphides unite to form the concentrated matte.

D. H. Browne made observations on this subject. Under modern conditions this preliminary concentration of the crude matte is unnecessary, and they can be refined directly; but, in any case, the coarse matte can be concentrated or refined by an oxidizing roasting followed by a reducing smelting as just indicated for the treatment of poor mattes, or the crude matte is enriched as in the case of copper by melting it in a cupola and running it into a ladle whence it is charged into a bessemer converter, or the molten matte may be run into ladles directly from the blast-furnace, and thence on to the converter. The converter is made like those used in purifying copper, and they have a quartz lining. During the blow, which may occupy 25 to 90 minutes dependent on the proportion of iron, quartz sand is added to slag the iron, and to prevent undue corrosion of the lining. When the proportion of iron has been reduced to about 0-5 per cent., the blast is stopped, and the matte is poured. The refining of the matte by a blast of iron was suggested by T. Gibb and C. Gelstharp in 1870. Attempts to carry the refining process further have not been successful because up to a certain stage the nickel is more easily oxidized than sulphur, and because the heat liberated by the oxidation of the sulphur is not sufficient to keep the matte in a fluid condition.

The proportion of sulphur is much less than is needed to form sulphides with the metals present. The copper is probably present as cuprous sulphide, and the nickel and iron partly as sulphides and partly as solutions of the metals in the mixed sulphides. The refining process at Port Colborne was described by R. L. Peek, and there are several modifications - e.g., by J. Savelsberg.