THE GOLDSCHMIDT REACTION. 



Bv H. STANLEY RKDGROVI-:. r>.Sc. (Lond.). F.C.S. 



1. Scip;ntific Aspects of the Goldschmiht 

 Reaction. 



It was ill 1N9S that Dr. Hans Goldschniiilt 

 announced that he had succeeded in reducing the 

 o.xides of many metals very conveniently by the aid 

 of aluminium^ That aluminium should pr<)\e a 

 most powerful reducing agent, readih' reacting with 

 the oxides of other metals with the [)roduction of 

 much heat, is a result that might ver\- well be 

 expected from the fact that aluminium has a greater 

 heat of combustion, and therefore possesses a greater 

 affinit\" for oxygen than the 

 majority of other metals. 

 But like certain other sub- 

 stances which, capable of 

 reacting e xotherm ically 

 (i.e., with the e^•olution of 

 heat) with one another. 

 will not do so until a cer- 

 tain amount of energy has 

 been supplied to them from 

 without, aluminium and 

 metallic oxides will react 

 with one another only at 

 high temperatures. Earlier 

 attempts, however, to 

 bring about the reduction 

 of metallic oxides In' 

 aluminium, by heating 

 mixtures of these in- 

 gredients, proved \'ery 

 unsatisfactory; for 

 either no reaction took 

 place, not sufficient heat 



rr T 



^yyyyyyM'/y:i^y/y/)'y>%y///^^^ 



to start it being sup- 

 plied, or else it occurred 

 with explosive \iolence. 

 Dr. Goldschmidt over- 

 came the difficult\ by an ingenious device. 



FicruH 1. 

 \pparLitus employed by Weston and Fllis for carryin 

 "therniitic " reactions ;/; vacuo. 



He found 



The external heating necessary to start the reaction 

 may in some cases be supplied by means of a strip 

 (if binning magnesium ; or better, a fuse composed 

 ri- of aluminium powder and liarium 



peroxide (which can be ignited by a 

 flaming vesta) may be employed. 



I'rom the chemical standpoint the 

 reaction is one of the simplest, thus, 

 in the case of iron oxide it may be 

 represented by the following equa- 

 tion, — 



FeoOa + 2 Al = AloO, + -' Fe, 

 — and similar equations can be con- 

 structed to represent the reactions in 

 the case of other metallic oxides. The 

 heat generated in the reaction between 

 aluminium and iron oxide is sufficient 

 to fuse both the alumina and the 

 metallic iron produced, the whole 

 contents of the crucible in which the 

 reaction is carried out becoming fluid. 

 The metal, owing to its density, 

 sinks to the bottom. The temperature 

 reached is second only to that of the 

 electric furnace, being estimated to be 

 about 3.000"C. The reaction can 

 be \er\- readily carried out, no 

 il)liaratus be\'ond a crucible of 

 h i g h 1 \' refractory 

 material being required. 

 Care, however, should 

 be taken to protect the 

 crucible in case it 

 should be cracked b\- 

 the heat of the reaction, 

 and its molten contents 

 run out. 



That substances 



out 



apparenth' so inert as aluminium and iron oxide 



that intimate mixtures of metallic oxides and powdered should be capable, once interaction between 



or granulated aluminium, if heated strongly at one them has been started, of producing suffi- 



point, react with the production of intense heat, the cient heat, not only to continue their own com- 



temperature produced being far greater than that bustion, but to liquefy iron and even more difficultly 



required to initiate the reaction. Consequently the fusible metals, may seem an extraordinary fact. But 



reaction spreads throughout the whole mass, as a it is (juite analogous to the behaviour of a rock nicely 



matter of fact doing so ver\- rapidly, the time of balanced at the top of a hill. Leave it alone, and 



reaction, which is about half to one minute in the what could seem more destitute of energy : give it 



case of iron oxide and aluminium, not appreciably but a gentle push — and who shall stay its course ? 



depending upon the cpiantitv of material employed. r>y using other oxides in place of iron oxide 



' H. Goldschmidt: " Ueber ein neues Verfahren /ur Darstellung von Mctallen und Legirungen niittelst .-Muniiniunis" .Aniujlcii 



dcr Chcinic, (1898), Vol. 301., pp. 19 et seq.. and H. Goldschmidt and C. Vautin : " .'\luminiuin as a Heating and Reducing 



Agent" Journal of the Society of Chcimcal Industry, (1898), Vol. 17, pp. 543, 649. 



249 



