Demonstration 17
Thermite reaction.
Soon after metallic aluminium was first isolated, both Sainte-Claire Deville and Wöhler noted its property when mixed in the form of powder or granules with the oxides of some of the metals, of reducing them with the evolution of sufficient heat to melt the metal and the alumina produced [1].
On this was based the process applied by Dr Hans Goldschmidt of Essen, who in 1898 was developing commercially the promising work of Claude Vauntin and Hugh Picard in London, following the change in the economics of aluminium supply which had taken place when production by the Héroult and Hall electric furnace method superseded the expensive sodium reduction process; prior to that time the use of aluminium as a reducing agent was little more than a scientific curiosity [2]. Goldschmidt has obtained either the metal, or an alloy of the metal, with aluminium, from the oxides of chromium, manganese, iron, copper, titanium, boron, tungsten, molybdenum, nickel, cobalt, zirconium, vanadium, niobium, tantalum, cerium, thorium, barium, calcium, sodium, potassium, lead and tin. He said:
In a thermite reaction, a metallic compound is reduced by one of several metals or metallic alloys in such a way that when the mixture is ignited at one place, the reaction continues of its own accord, so that under complete oxidation of the reducing element, a fluid slag is formed, while the reduced metal is obtained as a homogeneous uniform regulus; if the oxide is used in excess, the reduced metal is free, or practically free, from the element used as a reducing agent.
Goldschmidt’s significant innovations comprised the means of starting the reaction by a fuse, instead of heating the mixture until ignition took place, and the various controlling procedures which made the process practicable on a plant scale. He also arranged ingenious adaptations to produce molten iron or steel for the welding and repair of rails and machine parts and for many years these have been familiar on tramways and railway systems and in rolling mills.
This is the first of two solid/solid reactions, see also experiment 18a. This reaction, uses powdered aluminium metal to reduce metallic oxides and is especially useful for the reduction of those metal oxides which are difficult to reduce such as titanium and molybdenum. It is also known as the Goldschmidt process, and the Aluminothermic process. This is a highly exothermic reaction and the metal emerges in its molten state often very much above its melting point. A mixture of finely divided aluminium with ferric oxide (or the oxide of some other metal), approximating to 2Al+Fe2O3, is sold under the registered name Thermit (no final ‘e’), and is used in joining or welding iron and steel rails, pipes, etc. Sections of railway lines are usually welded in this way and the reaction is mostly associated with the reduction of iron oxide as in our demonstration. Like the first part of the next demonstration this reaction takes place between solids and is, consequently, difficult to start.
Two equivalents of finely divided aluminium and one of iron oxide are mixed together and ignited. The reaction takes place as follows:
Fe2O3 + 2 Al ==> Al2O3 + 2 Fe + 848.54 kJ (17.1)
Preparation. We can no longer obtain fire clay crucibles (we used to use Morgan No.1) instead we now use two ceramic plant pots one stacked inside the other. The inner pot is 65 mm diameter and 68 mm high and the outer one larger so that the smaller one fits snugly into it, with the proviso that the holes are close together. The outer pot in necessary because the inner one almost always breaks up due to the intense heat. These are attached to the stand using a ring clamp and boss. A small piece of filter paper is placed into the bottom of the inner pot to cover the hole before filling ca. 2/3 of its volume with the Thermite mixture which must be thoroughly dry. The stand is placed such that the ‘crucible’ is centred over and about 50 cm above a metal box containing sand to a depth of at least 8 cm. The surface of the sand is made slightly concave, so that the middle is lower than the edges. The whole being placed on two or three heat resistant mats. Modern, asbestos free, mats are composed of a ceramic material held by a cellulose binder and the molten metal produced in this reaction can burn into or through them. Safety shields must be set around this experiment. The rest of the bench top and the surrounding floor is best protected by covering with sheets of thin medium or high density wood composite board, e.g., ‘Hardboard’. Asbestos free fire blankets are made of glass fibre and are of little use.
Demonstration. Check that the crucible is centred over the sand box and that the sand slopes down from the edges of the box, ensure also that the safety screens are securely in position. Reactions between solids are difficult to activate because the contact between the reactants is poor. There are various methods of generating sufficient heat to start this reaction (see below). We use a sparkler of the type used by the professionals who weld railway lines by the Thermit process. The sparkler is placed vertically into the mixture leaving some of the coating above the surface and, using a blow torch, this part of the sparkler is ignited. Stand back immediately. After a few seconds sparks, flame and smoke will be ejected from the crucible and after a short delay molten iron will run from the hole in the bottom of the crucible onto the sand. The hot mass of iron which is coated with ‘slag’ from the molten Al2O3 and fused sand should be left to cool to dull red heat, when it can safely be cooled rapidly by being placed in a container of water. Warning: hydrogen can be produced at this stage if the regulus is too hot. When cool the iron core can be separated from the slag using a hammer, and the iron identified with a magnet.
Alternative methods of starting the reaction.
i. Potassium chlorate and sugar (grind separately), place on top of the Thermite mixture, make a small hollow in the top of the heap and add a few drops of conc. (18M) sulphuric acid.
ii. Place on top of the Thermite mixture some 20g of potassium manganate(VII) (permanganate), there must be some powder or fine crystals present, make a hollow in the top and, when ready, add about 5 cm3 of glycerol.
iii. Similarly we can use a mixture of aluminium metal and barium peroxide and employ a magnesium ribbon to ignite the mixture [3].
We use the chrome(III) oxide Cr2O3 from experiment 13 in a small-scale version of this reaction (by igniting a small heap of the dark green oxide and aluminium in a sand tray) [4]. Cr2O3 and other mixtures of aluminium and metal oxides have been described by Roebuck [5] including MnO2, Mn3O4, CuO, NiO, CoO, and Fe3O4. However, CuO + Al, and Mn3O4 + Al have been said to have reacted explosively, Crellin [6]. Shakhashiri [7] urges that only Fe2O3 + Al be used. A demonstration on preparing silicon by Thermit process has been described together with the directions for making plaster of Paris crucibles [8].
A great variety of practical application had been proposed for the thermite reaction. These included a successful slow-burning thermit which was devised during the Second World War for cartridges capable of heating cans of soup and other foods required on active service.
Safety. This is a hazardous procedure, but is a safe demonstration in experienced hands. However, there must always be a first time, and it is to those persons that we offer the following advice. Read the safety recommendations of say reference 7 below.
First perform this experiment outdoors a couple of times, preferably use a table as you would indoors. There are very few demonstrations that can safely be performed for the first time in front of an audience and, in this case, we recommend that you practice at least twice more indoors before regarding yourself as sufficiently experienced and competent. It is always a good idea to get someone else to work with you as an equal partner, in that you should both be equally familiar with the procedure, and both sense the burden of responsibility. In these circumstances the partners should assume individual and collective responsibility and so will ‘watch out for each other’. For example a teacher might involve his technician in this way.
It is important that the Thermite mixture is perfectly dry otherwise extremely hot material might be violently ejected from the crucible. Dry the components separately at ca. 125C. Do not use heat to dry the mixture use a desiccant.
Metallic iron is produced at a temperature significantly above its melting point at 1535C. Once the reaction has started it is almost impossible to stop it. A CO2 fire extinguisher should to hand; water should not be used because potentially explosive hydrogen can be produced. No one should be closer to the reaction than two metres.
References
1. Thorpe’s Dictionary of Applied Chemistry, 4th ed., vol. XI, London - New York - Toronto, Longmans, Green and Co., 1954, pp. 562-565.
2. H. Goldschmidt, German Pat., D.R.P. 96317, 1895; Z. Elektrochem., 1898, 4, 494; Ibid., 1899, 6, 53; Electrochem. Metall. Ind., 1908, 6, 360; Stahl Eisen, 1898, 18, 408; H. Goldschmidt and C. Vautin, J. Soc. Chem. Ind., 1898, 17, 543.
3. J.W. Mellor, A Comprehensive Treatise on Inorganic and Theoretical Chemistry, vol. 5, new impr., London, Longman, Green and Co., 1956, p. 218.
4. Tested Demonstrations in Chemistry, ed. G.L. Gilbert, et al., vol. 1, Granville, OH, Denison University, 1994, p. H-9.
5. P.J. Roebuck, Educ. in Chem.(Britain), 1979, 16, 178.
6. J.R. Crellin, Educ. in Chem. (Britain), 1980, 17, 93.
7. B.Z. Shakhashiri, Chemical Demonstrations, A Handbook for Teachers of Chemistry, vol. 1, Madison, The University of Wisconsin Press, 1983, p. 88.
8. G. Fowles, Lecture Experiments in Chemistry, 3-rd ed., London, G.Bell & Sons, Ltd., 1947, pp. 285-287.