Page:Encyclopædia Britannica, Ninth Edition, v. 13.djvu/296

 280 IRON chosen rather than the older combining number 28 in accordance with Dulong and Petit s law. Besides these, however, indications of the existence of a lower oxide Fe 2 have been obtained by Lowthian Bell (Chemical Phenomena of Iron Smelting, p. 85) ; for by partially reducing the higher oxides by carbon oxide at temperatures near to 420 C., a mixture of metallic iron, unreduced oxide, and free carbon results, from which the iron can be dissolved out by digestion with water and iodine in closed vessels, after which the relationship between the undissolved iron and the oxygen present is very close to that indicated by the formula Fe 2 O. On the other hand, derivatives of an oxide higher than Fe 2 O 3 exist, compounds known as ferrates being formed by heating iron with nitre, and in other ways, the .composition of which may be expressed by regarding them as containing iron trioxide united to other metallic oxides, e.g., potassium ferrate, K 2 O, FeO 3 (just as potassium sulphate may be regarded as an analogous compound con taining sulphur trioxide, K 2 O, SO 3 ). Neither the hypothe tical ferric anhydride (or iron trioxide, Fe0 3 ) nor any other oxide intermediate between it and Fe 2 3 has as yet been isolated, although iron disulphide, FeS 2, is well known. Of these oxides, two, viz., FeO and Fe 2 O 3, correspond to stable well-defined classes of salts conveniently referred to as the ferrous and/erne salts respectively, 56 parts of iron replacing 2 parts of hydrogen in an acid to form the cor responding ferrous salt, and replacing 3 parts to form a ferric salt. The heat of formation of all oxides up to Fe 3 O 4 appears to be about uniform, viz., near to 66,000 gramme degrees per 16 grammes of oxygen combined ; but that of the formation of Fe 2 O 3 is sensibly less, so that when the latter oxide is reduced the rate of reduction is much more rapid prior to the removal of one-ninth of the oxygen present and consequent formation of Fe 3 O 4 than it is subsequently, in accordance with the general law which appears to exist connecting the rate at which reduction goes on with the development of heat during the chemical change (Alder Wright and Rennie, Chem. Soc. Journal, 1880 [Transactions], p. 757). Probably it is in consequence of this that Fe.,O 3 breaks up at an intense white heat into oxygen and Fe 3 O 4, and that when iron is burnt in oxygen so that the temperature is very high Fe 3 4 is formed and not Fe 2 3 ; just as higher oxides of manganese than the corresponding Mn 3 O 4 break up on strong ignition into oxygen and Mn. s O 4. On the other hand, just as oxides of manganese lower than Mn 3 4, and also that oxide itself, take up oxygen forming higher oxides on heating in the air to moderate temperatures, so Fe 3 O 4 can be oxidized to Fe 2 O 3 by direct addition of oxygen taken up in the s ame way ; it is noticeable, however, that, whilst Fe 3 O 4, possess ing a certain kind of physical structure, will thus oxidize to Fe 2 3 on being exposed to ordinary atmospheric influences, yet when other kinds of physical structure are possessed (producible by special modes of formation) its tendency to oxidize further even in moist city air becomes inappreci able. It is remarkable that, whilst iron ores which mainly contain the oxide Fe 3 O 4 are highly magnetic in character (the loadstone being one variety of this class of minerals, and the generic names &quot; magnetic iron ore &quot; and &quot; magnetic oxide of iron &quot; being derived from this fact), the other iron compounds found in nature are far less marked in regard of their possession of this quality, one particular sulphide of iron exeepted, termed magnetic pyrites in consequence, and indi cated by the formula Fe 7 S 8, or possibly Fe 8 S 9. Thus the following values were found by Thicker as the relative effects of equal vol umes of soft iron, loadstone, specular iron ore, and brown haematite on a given magnet under similar conditions : Soft iron 100- Native magnetic oxide 40 227 Specular ore 0-533 Brown hcematite 0-071 Oxides of iron of all classes are readily acted upon by reducing agents (especially hydrogen, carbon, oxide, and free carbon and silicon) in such a fashion as to cause the transference of the oxygen of the oxide to the reducing agent, a lower oxide of iron and finally metallic iron being set free. Jn accordance with the general rules obtaining in such cases (Alder AY right and Itennie, loc. cit.), the rate of reduction of ferric oxide of given physical character is less, ceetcris paribus, when a reducing agent is employed which evolves less heat in uniting with oxygen than when one is used evolving more heat ; so that a reduction by hydrogen with formation of water vapour goes on more slowly under constant conditions than reduction by carbon oxide forming carbon dioxide, whilst the temperature requisite to cause reduction to be brought about to a just measurable extent (temperature of initial action) is lower with carbon oxide than with hydrogen, and lower with hydrogen than with free carbon (Alder Wright and Luff, Ohcm. Soc, Journal, 1878 [Transactions], pp. 1, 504). The precise rate of reduction and temperature of initial action observed in any given case vary with the conditions of the experiment and also with the physical char acter of the iron oxide (see also Lowthian Bell, Chemical Phenomena of Iron tinicltiny). When ferric oxide is reduced by carbon oxide, a peculiar second ary change is brought about under certain conditions, which has been investigated by Lowthian Bell with the present writer s co operation (loc. cit.) ; this consists in the reaction of a lower oxide of iron (Fe. 2 ?) formed at a certain stage of the reduction on the carbon oxide forming a higher oxide of iron and setting free car bon ; 1 the higher oxide of iron is then again reduced by a fresh portion of carbon oxide, and so on in a cycle, so that after some time the quantity of free carbon deposited largely exceeds the total iron present. This peculiar action is also exhibited by oxides of nickel and cobalt, but apparently by those of no other metals ; it has a most remarkable influence upon the nature of the chemical changes ensuing in the process of smelting iron by the blast furnace ( 19), and is doubtless the chief source of the carbon con tained in pig iron thus produced ; it is also the main reaction taking place during the conversion of iron into steel by cementation ( 32). Ferrous carbonate differs from most of the other compounds of iron found in nature in being soluble in water, especially when excess of carbonic acid is also present, an &quot; acid carbonate &quot; being formed. Such water on exposure to air forms a rusty deposit of hydrated ferric oxide produced by the combination of the oxygen of the air with the ferrous oxide contained in the ferrous carbonate, the carbon dioxide originally combined therewith being set free. In certain localities large deposits of more or less pure hydrated ferric oxide are thus formed, constituting bog iron ores. &quot; The sulphides of iron partly correspond to the oxides. Thus the sulphides Fe 2 S, FeS, and Fe 2 S 3 exist ; besides these, the compound Fe g S has been described, whilst magnetic pyrites, Fe 7 S 8 (or Fe 8 S 9 ), and ordinary pyrites, FeS 2, and its allotropic or metameric modification mar- casite, constitute minerals of widespread occurrence, and of considerable value, mainly as sources of sulphur, secondarily on account of the iron they contain, and more especially with certain kinds of pyrites on account of the copper, silver, and gold sulphides intermixed therewith. It is to be noticed in connexion with pyrites that, by the action of reducing agents on solutions of iron compounds in presence of sulphates, a slow formation of crystalline FeS 2 often results ; thus many fossil plants and animals occur in various strata in which the deposition of pyrites by this means has produced a perfect cast or pseudomorph, so to speak, of the organism ; it is probable that the pyritous deposits of large magnitude which exist in various localities have been formed by these agencies, the soluble iron salt having been originally the carbonate. The chlorides of iron correspond to the ferrous and ferric series of salts, i.e., are indicated by the formula? FeCl 2 and FeCl 3 (or preferably Fe 2 Cl 4 and Fe 2 Cl (i ) respectively ; chlorides corresponding to Fe 2 O, Fe s O 4, FeS,, Arc., have not as yet been formed. The same remark applies to the salts of iron formed by the substitution of iron for hydrogen in all the acids of common occurrence. For the use of iron salts and other ferruginous compounds in the arts 1 According to Griiner (Comptes Rendus, 1871, 28), the reaction is SFeO + CO = Fe 3 4 + C, some metallic iron being always formed in addition to the ferrous oxide produced by the subsequent reduction of the Fe 3 4 , so that a fer ruginous carbon always results.