Page:Encyclopædia Britannica, Ninth Edition, v. 13.djvu/326

310 IRON to the other constituents ; the reaction is not sensible with hard coke at temperatures lower than 300, whilst at 400 and somewhat upwards it is not marked; at 500 and G00, however, it goes on pretty rapidly, the more so the less hard and dense the coke, charcoal acting much more readily under similar conditions than coke. Accordingly as the iron ore and the fuel gradually sink in the furnace and become hotter, they tend to affect the composition of the gas in opposite ways, the former decreasing the carbon oxide and increasing the carbon dioxide, and vice versa with the latter. The rate of reduc tion of iron oxide under constant circumstances is, however, a diminishing one, inasmuch as the reduced particles cover up the unreduced ones and prevent their being so readily acted on ; so that, whilst on descending into a hotter region the rate of reduction of the ore is at first increased owing to increase of temperature, by and by the rate of removal of oxygen as it sinks ceases to increase and ultimately diminishes. Long before anything like complete reduction is brought about, however, other changes are brought into play which greatly modify the actions. As soon as the iron ore is partially reduced, it begins to react on the carbon oxide in the way indicated by the equation Fe^Oy + CO-C + fWu ..... (4), setting free finely divided amorphous carbon in contact with it. 1 Again, as soon as metallic iron in a spongy form is pro duced, it reacts on the carbon dioxide, thus scFe + yCOs-FesOf + ^CO ..... (5); whilst very probably a parallel reaction takes place with lower oxides of iron not completely reduced to the metallic state, these actions being practically reciprocal to those in virtue of which carbon oxide reduces to ferric oxide, first to a lower oxide and then to metal. Yet again, when carbon and iron oxides are heated together, there takes place a change virtually reciprocal to that in virtue of which carbon is deposited from carbon oxide (equation 4), carbon oxide and dioxide gases being formed, and the iron oxide being more or less reduced in virtue of the reactions The ultimate result then is that before the ore and fuel have descended far they are subjected to a number of oppos ing forces so far as the ore is concerned, the carbon oxide in the gases surrounding it and the deposited carbon in con tact with it tend to remove oxygen by reactions 3, G, and 7, whilst the carbon dioxide in the gases and the reaction causing deposition of carbon from carbon oxide tend to re- oxidize it by reactions 4 and 5 : the fuel and carbon oxides in the gases on the other hand are analogously affected ; the reaction of the carbon dioxide on the fuel, 2, tends to gasify the latter (the action being more rapid with charcoal than with coke Lowthiau Bell, also Akennann), and that of the carbon oxide on the partly reduced iron ore setting free carbon, 4, to reverse this action. The actions of the iron and its oxide on carbon, and on carbon oxide and dioxide, also are opposed, some tending to increase the carbon oxide, 5 and 6, and some to decrease it, 4, and others to affect similarly the carbon dioxide, viz., 3 and 7 to increase it, and 5 to decrease it. In consequence, at any given level of the furnace a sort of compromise is arrived at amongst all these varied oxidizing and reducing influences, the net or resultant chemical action being that, whilst a portion of the hard coke of the fuel is gasified, and reciprocally a portion of finely divided amorphous carbon precipitated from the gases, the iron is partially but not wholly reduced. On the whole, then, as the ore sinks in the furnace, it 1 According to Griiuer (Comptes Rendus, 1871, 28) this reaction is and does not commence until the iron ore is deoxidized to some con siderable extent, at least on the outer surface of the lumps of ore. becomes hotter and hotter and more and more deoxidized, but owing to the oxidizing influences at work it does not part with all its oxygen until it has descended some considerable distance to a point where the temperature is about sufficient to fuse it ; at this stage the last portions of oxygen are removed, partly by the precipitated amorphous carbon, partly by the alkaline cyanides accumulating in the furnace, and the almost completely reduced metal melts, dissolving as much of the amorphous carbon in contact with it as it can take up under the circumstances; simultaneously the silicious and earthy matters present also fuse, forming cinder. The reducing influences at work here also cause the deoxidation of some of the silica present, whilst man ganese, phosphorus, and sulphur compounds, &c., are also more or less reduced and taken up by the fusing iron. When the proportion of fuel relatively to the burden is diminished, a larger amount of incompletely deoxidized ore reaches the hearth, the result of which is that, as the silicious and earthy matters fuse, they dissolve some of the iron oxide before it has time to become reduced by the deposited carbon, giving a ferruginous cinder, whilst this carbon is used up in completing the reduction more rapidly than would otherwise be the case ; the pig iron formed is less highly carbonized than before, becoming white instead of grey, partly owing to the diminution in the quantity of dissolved carbon, and partly because the temperature of the hearth is lowered, and there is less time for graphite to separate in cooling, The formation of alkaline cyanides and their reaction on the imperfectly reduced iron oxide is brought about as follows : in the upper part of the furnace a crust of alkaline carbonates, &c. , carried up as fume by the escaping gases ( 18), is deposited on the surface of the materials, and so is brought down again to the hearth, where the nitrogen of the blast and carbon act on it conjointly, forming (for potassium carbonate) potassium cyanide, thus K a C0 8 + N s + 4C = 2KCN + SCO . The exact nature of the reaction of potassium cyanide on the imperfectly reduced iron oxide with which it finds itself in contact is not known, but it is probable that potassium oxide and iron cyanide are formed, the latter becoming decomposed into iron, car bon, and free nitrogen, and the former being carried away by the escaping gases and deposited as potassium carbonate in the upper part of the furnace, so that where the cyanide is formed (mainly at or near the tuyere level) there is an evolution of carbon oxide and a disappearance of nitrogen, whilst a little higher up there is a re- evolution of nitrogen ; that is, whilst at the tuyere level and there abouts the carbon and oxygen in the gases are raised, relatively to the nitrogen, considerably above the amount due simply to the blast becoming transformed into carbon oxide and nitrogen, a little higher up the amounts of carbon and oxygen appear to diminish relatively to the nitrogen ; not that they actually do diminish in quantity, but that the evolution of nitrogen from the cyanide decomposition causes their amounts to be lessened relatively to the total nitrogen. Thus the following numbers are calculated from some of Lowthian Bell s observations with an 80 foot furnace using coke and calcined Cleveland ironstone, the gases being obtained by drilling holes through the furnace wall at the different levels, and collecting the issuing gas ; the amount of carbon in the gases is manifestly greater at the tuyere than that due to the blast ; for some feet it apparently diminishes owing to the cyanide reaction, and then remains almost constant till near the top, where it increases from the expulsion of carbon dioxide from the iiux. The oxygen again is considerably in excess of that due to the blast at the tuyeres, but at a somewhat higher level it apparently decreases, whilst higher up still it increases again owing to the reduction of the ferric oxide and the evolution of carbon dioxide from the limestone. Composition by Weight of Gases at different Furnace Levels. Height above tuyere 1" 1-2 37-6 fll-2 6 12 20 37 50 1-2 34-8 G4-0 60 3-5 33-2 63-3 765 Blast if wholly burnt To CO. To C0 2 . Carbon dioxide Carbon oxide trace 37 1 629 0-8 35-9 63-3 1-2 34-9 03-9 16 34-8 63-C 7-9 33-0 59-1 84-4 656 29-2 70-8 Carbon and Oxygen calculated per 100 of Nitrogen. Carbon 26-8 36-5 25-2 3,3-7 246 33-3 23-9 32 6 24-1 33-1 23-8 32-4 240 33-9 27-5 41 6 22-5 30-0 11-3 300 Oxygen