Advanced Search
RSS LinkedIn Twitter

Journal Archive

Platinum Metals Rev., 1992, 36, (3), 146

Refining Secondary Iridium by an Oxidative Method

  • By A.V. Yermakov
  • V.M. Koltygin
  • E.V. Fatyushina
  • Laboratory of Strength, Ural State University, Ekaterinburg, Russia

The majority of investigators reporting on the mechanical properties of iridium have noted its high sensitivity to impurities. The elements which may have caused embrittlement are carbon, oxygen, and the refractory metals tungsten and rhenium. The segregation of these impurities to the grain boundaries in pure iridium has been considered to be the cause of brittleness in samples which failed after less than 5 per cent elongation at room temperature (1). The reduction in metallic impurities, and in carbon and oxygen, to a level of parts per million, allows iridium to be successfully used at temperatures of 1100 K and higher (2).

Iridium crucibles are subjected to repeated thermal cycles for long periods of time at high temperatures. These large iridium crucibles are used to grow single crystals of gallium-gadolinium-garnet and aluminium-yttrium-garnet. The resulting scrap material may be contaminated with platinum, palladium, rhodium, iron, tungsten, molybdenum, magnesium, nickel, silicon, calcium, copper, oxygen, carbon and oxides of aluminium, yttrium, gadolinium, gallium and zirconium. The scrap is likely to contain 95 to 98 weight per cent iridium. The conventional method of producing high purity iridium by chemical refining is extremely labour intensive, and results in significant loss of the metal (3).

A number of alternative methods of purification are known, and one of them is considered here. This process involves oxidation of the impurities during air melting of the scrap iridium in a periclase crucible (periclase contains, in weight per cent, 96 MgO, 1 A12O3, 1.2 CaO, 1 Fe2O3 and 0.8 SiO2). During this process the iridium exhibits extreme inactivity at the high temperatures used, 2715 to 3000 K (4). The reactions of the contaminating materials can be classified into two groups according to their characteristic interactions with oxygen in the air at the melting point of iridium, and at even higher temperatures:

  • Elements which form stable oxides: aluminium, yttrium, gadolinium, gallium, iron, tungsten, carbon, molybdenum, niobium, magnesium, silicon and zirconium.

  • Elements having oxides which dissociate below the melting temperature of iridium, namely: platinum, palladium, rhodium, nickel and copper.

Calculation of the Concentration Equilibrium of Impurity Elements

The possibility of removing the first group of elements from iridium was calculated from the Free Energy of these elements at 3000 K and pressure of 102 Pa, and from the resultant equilibrium concentrations of the impurities and the stable oxides. The equilibrium concentrations of the oxides of these elements with liquid iridium was determined from the equilibrium concentration of the impurities, and the final analysis of the liquid iridium. Melting the iridium in a periclase crucible formed a slag based on periclase which allowed better removal of most of the impurities. The calculations were performed on systems consisting of “iridium with impurities-oxygen in air”, at temperatures of 2800 to 3000 K, approaching those of the ideal solutions (5). During the calculations, consideration was given to the condensation of the metal and slag phases and also to the migration, in the gas phase, of volatile oxides which have high partial pressures. Atomic adsorption analysis data show that concentrations of oxide impurities in the slag, which did not enter the periclase, were not more than 0.1 weight per cent. Partial pressures of gas forming oxides in the given temperature range have been studied previously (6). Calculated values of the maximum equilibrium concentrations of the impurities in iridium after induction-oxidation melting are given in the Table.

Calculated and Experimental Data of the Impurity Concentrations of Elements in Iridium Scrap after Induction-Oxidation Melting

ElementInitial concentration, at. %Partial pressure of oxides at 3000 K, × 102 PaEquilibrium concentration of impurities in liquid iridium, at.%, in systemsExperimental contents of impurities, at.%
Al* 0.2 0.813 1.6 × 10-3 1.5 × 10-2 4 × 10-4
y* 0.05 1.093 4.3 × 10-5 n.r. 2.6 × 10-6
Gd 0.15 0.533 2.9 × 10-5 n.r. 2.4 × 10-5
Zr* 0.05 0.187 2.6 × 10-5 n.u. 4.6 × 10-5
Si 0.01 289.3 1.5 × 10-3 4 × 10-2 8.9 × 10-4
Nb 0.05 19.5 1.7 × 10-3 n.u. 1.9 × 10-5
Fe 0.8 169.3 8.6 × 10-2 2.2 × 10-1 1.2 × 10-1
Ca 0.01 0.2 4 × 10-3 n.r. 3.4 × 10-5
W 1.5 45322 3.7 × 10-4 n.u. 4.3 × 10-4
Mo 0.3 19995 2 × 10-3 n.u. 1.5 × 10-4
Ga* 0.1 2266 1.5 × 10-2 5 × 10-7 7 × 10-3
C 0.1 1013 9.1 × 10-5 n.u. n.o.
O 0.5   4.9 × 10-1 n.u. n.o.
Cu 0.06 58.65 n.u. 2.3 × 10-3 2.2 × 10 -3
Ni 0.03 738.5 n.u. 8.5 × 10-3 3.2 × 10-3
Pd 0.01 n.u. 2.3 × 10-2 2.2 × 10-2
Mg 0.01 48 2.7 × 10-1 4 × 10-4 7 × 10-4

Al. Y. Gd. Zr. Ga-additive elements found in charge as oxides

– impurities could not be removed

– contents of elements were not determined

– equilibrium contents of additives were not calculated

In addition, iridium purification could occur by evaporation of some impurities from the melt. These processes are limited by the fact that melting takes place at atmospheric pressure, but contaminants from both groups with high partial vapour pressures were removed from the iridium.

In previous work the systems “iridium with impurities-inert gas (argon)” were studied at temperatures of 2000 to 4000 K and at pressures of 9.8 × 10-2 to 1.3 × 10-8 MPa, and the evaporation of elements such as platinum, rhodium, aluminium, magnesium, tungsten, iron, nickel, copper, silicon, zirconium and carbon were described (7). Calculations were performed on the basis of the “maximum entropy” principle: entropies of the gaseous phase and of the components in the condensed state. The low concentrations of impurities allow calculation of ideal solutions. The results of the calculations are given in the above Table.

Experimental Studies of Impurity Behaviour

In order to compare theoretical and experimental data, a series of oxidation-induction melting experiments were performed on iridium scrap containing the impurities listed in the Table. Melting was carried out in air using a periclase crucible with porosity of less than 4 to 5 per cent. The iridium was maintained in the liquid state for a period of 30 minutes, at a temperature of around 3000 K. Before the metal was cooled in the crucible, 20 to 30 g samples of molten iridium were taken for control measurements. The compositions of the impurities were established by mass spectroscopy. The selected regimes allowed measurements to be made of impurities present in the iridium at concentrations of the order of 10-7–10-8 atomic per cent. The experimental data are also presented in the Table.

The experimental concentration values for aluminium, yttrium, gadolinium, zirconium, silicon, iron, molybdenum, gallium, copper, nickel, palladium and magnesium, agree well with the calculated data. On the basis of these results, it is possible to conclude that iridium refining can be carried out by the following methods:

  • The oxidation of impurities which form stable oxides during the induction-oxidation melting process. These, together with the oxides in the furnace charge, then enter into the slag. The removal of aluminium, yttrium, gadolinium, zirconium, silicon, niobium, iron and calcium from the iridium was carried out by this method.

  • The formation of oxides of the contaminants which have high partial pressure, such as tungsten, molybdenum, gallium, carbon, and which evaporated during melting.

  • The removal of impurities such as copper, nickel, palladium and magnesium which occurred due to evaporation because they have higher partial pressures than iridium.

For calcium and niobium, the level of purity obtained was higher than that calculated. This may be due to the formation of chemically stable oxides with MgO, resulting in a sharp fall in CaO and NbO activities in the slag and a decrease of calcium and niobium concentrations in the melt. The final concentration of tungsten was higher than the calculated value. This result shows that tungsten removal is dependent on the kinetics of the reaction for its removal from the melt. When air was introduced into the iridium melt the tungsten formed a volatile oxide and its removal was increased (8).

Discussion of the Results

The agreement of the calculated data with figures obtained experimentally showed that the method of calculation used, and the assumptions made, were correct. Effective removal of contaminants from iridium results from better contact of the melt with the periclase crucible, thus binding part of the oxides, and intensive mixing of the melt with the help of an electromagnetic field. However, during induction-oxidation melting platinum and rhodium, which have properties similar to those of iridium, were not removed from the iridium. Also only a partial reduction in the concentration of the impurities of iron, gallium, palladium, oxygen, copper and nickel occurred. These elements can be removed more effectively from iridium by electron-beam melting, after which the concentrations of platinum and rhodium decrease to 10-3–10-4 atomic per cent, and residual impurities to 10-4–10-5 atomic per cent. The metal impurities which are removed in this way from iridium scrap have been picked-up during single crystal growth in crucibles and from iron-based apparatus used during the formation of these large crucibles.

After induction-oxidation melting iridium can be worked. The results of compression studies carried out at room temperature are given in Figure 1.

Fig. 1

Iridium deformation curves following compression at room temperature; specimens 1, 2 and 3 were cut from single crystals but specimen 4 was electron beam melted material. Specimen dimensions: 4 x 3 x 3 mm

Iridium deformation curves following compression at room temperature; specimens 1, 2 and 3 were cut from single crystals but specimen 4 was electron beam melted material. Specimen dimensions: 4 x 3 x 3 mm

Samples cut from a coarse grained ingot produced by induction melting were also investigated. This material was cold worked into single crystals that were found to be strong, which may be due in some way to the increased concentration of elements such as oxygen and iron in the iridium.



  1. 1
    B. L. Mordike and C. A. Brookes, Platinum Metals Rev., 1960, 4, ( 3 ), 94
  2. 2
    J. R. Handky, Platinum Metals Rev., 1986, 30, ( 1 ), 12
  3. 3
    B. Heshmatpour and R. L. Helstand, U.S. Patent 4,406, 693 ; 1983
  4. 4
    A. V. Yermakov,, A. I. Timofeev,, A. M. Orlov,, V. V. Shishkov and V. M. Koltygin, Vysokochistye Veshchestva, 1989, ( 4 ), 8
  5. 5
    O. Kubaschewski and C. B. Alcock, “ Metallurgical Thermochemistry ”, 5th Edition, Pergamon Press, London, 1979
  6. 6
    I. S Kulikov, “ Termodinamika Oksidov (Handbook) ”, Metallurgiya, Moscow, 1986
  7. 7
    N. A. Vatolin,, E. K. Moiseev,, P. A. Matyukhin,, A. I. Timofeev and A. V. Yermakov, Izv. Akad. Nauk SSSR, Met., 1986, ( 2 ), 50
  8. 8
    A. V. Yermakov,, V. M. Koltygin and V. M. Tereshchenko, unpublished work

Read more from this issue »