Platinum Metals Rev., 1987, 31, (2), 74
Some Ternary and Higher Order Platinum Group Metal Alloys
A Preliminary Review and Assessment of their Constitution and Properties
Platinum-based alloys find many applications in both high and low temperature industrial environments and are particularly suited to operation under corrosive aqueous and high temperature gaseous conditions. A wide variety of alloys exist which have been specifically designed for these purposes and in many cases they contain one or more non-platinum group metal. Although this situation is quite acceptable, there remains a fundamental need for systematic investigation of the basic properties and constitution of platinum group metal alloys. Such a foundation in materials technology generally leads to an understanding of materials behaviour and provides guidance in designing new alloys with improved properties for existing and future applications. This review features the work which has been done on ternary and higher order platinum group metal alloys and provides access to important data.
In July 1857 a patent was granted to Jules Henry Debray for “Improvements in the Manufacture or Reduction of Platinum” (1), and the English rights of this important patent were immediately purchased by George Matthey. It was claimed within his specification that the method developed for processing platinum ores was of particular advantage, in that alloys of platinum and the different metals found within its ores were produced which possessed properties more advantageous than pure platinum. Indeed, in 1859, H. Sainte-Claire Deville and J. H. Debray declared that they had solved the problem of preparing a platinum-iridium-rhodium ternary alloy with excellent ductility for fabrication (2). The weight percentages of rhodium and iridium specified by Deville and Debray were 5 and 19.6, respectively. The constitution of the platinum-rhodium-iridium system has not been well established and only recently has there been any attempt to characterise the structure and properties of selected ternary alloys (3).
By the mid-19th century it had been recognised that the properties of platinum could be improved by selective alloying with other platinum group elements, so extending the already growing industrial applications of platinum.
The constitution and properties of binary alloys of the platinum group metals have been thoroughly investigated and equilibrium phase diagrams exist on all but a few systems. However, little has been done to characterise the 20 possible ternary systems. Published information which is available on these systems is highlighted in Table I by a blue tint, while the type of data available on both these ternary and higher order systems is presented with references to the literature in Table II. Clearly, the platinum-palladium-rhodium system has received the most detailed investigation primarily because of the major industrial importance of binary platinum-rhodium alloys, for example, in catalysis and as a constructional material for glass processing equipment. The latter use has been considered in the first article in this issue of Platinum Metals Review.
|Possible Ternary Systems of the Platinum Group Metals|
System References and Data Type
The Platinum-Palladium-Rhodium System
Alloy compositions within the platinum-palladium-rhodium ternary system which have been reported in the literature are given in Table III. The range of properties which have been examined are shown in Table II. For the purpose of reviewing this system, specific areas of property characteristics will be highlighted although for more detailed analysis of the data the references cited should be examined.
Alloys Investigated in the Platinum-Palladium-Rhodium System
Compositions in weight per cent
Two of the three binary systems, platinum-rhodium (26, 27) and platinum-palladium (27, 28), which make up the ternary system, show complete solid solubility throughout the range of composition. Evidence for solid state miscibility gaps in all three systems does exist but it is in the rhodium-palladium system that this effect is most pronounced (29). No ternary phase diagram exists for this system although it undoubtedly shows complete solid solution characteristics over the whole range of alloy compositions.
Room Temperature Tensile Properties
The tensile strength (σ ) data presented in Table IV have been compiled from (4, 5, 6, 10, 30). Young’s modulus values (E) for binary alloys were obtained from (30, 31) and values for ternary alloys calculated assuming linear variation with atomic per cent throughout the ternary section. From both tensile strength and Young’s modulus the normalised tensile strengths (σ /E) have been calculated and presented as a series of profiles for three pseudo-binary platinum(palladium)-rhodium alloys with rhodium contents of 5, 10 and 20 weight per cent. These are shown in Figure 1.
In each of the pseudo-binary sections the terminal binary palladium-rhodium alloys exhibit the highest normalised tensile strength. It can also be seen that the tensile strength of platinum-rhodium-palladium alloys increases with palladium content. Apart from the strengths of the palladium-rhodium binary alloys, maximum strengthening in the ternary alloys occurs at a composition of 40 weight per cent palladium (approximately 50 atomic per cent in each case). Similarly, and as might be expected, an increase in rhodium content increases the strength of all the alloys.
The strength profiles obtained by plotting normalised stress against composition highlight non-ideal behaviour in alloys containing in excess of 40 weight per cent palladium. Tensile strength profiles presented do not reflect this characteristic, and indeed show binary palladium-rhodium alloys with rhodium contents of 5 and 10 weight per cent with tensile strengths below those of the equivalent binary platinum-rhodium alloys (6).
The most systematic investigation of the hardness characteristics of platinum-palladium-rhodium alloys has been carried out by Nemilov, Rudnitski and Polyakova (6). A full range of alloys within the ternary system was prepared from elemental sponge material containing less than 0.01 per cent impurities. Alloys were not chemically analysed as loss of material on melting accounted for less than 0.3 per cent. Effective homogenisation of the alloys could only be achieved after a few hours at 1300 to 1450°C. Brinell hardness measurements were taken on alloys in the annealed condition using a 10 mm diameter ball and a 250 kg load. Hardness profiles through pseudo-binary sections of the platinum-rhodium-palladium system with rhodium contents varying from 10 to 80 weight per cent are presented in Figure 2. High hardness is observed in alloys containing low percentages of platinum and relatively high rhodium contents.
Although under the conditions of material preparation and testing, the true low temperature equilibrium structure of low-platinum-rhodium-palladium alloys would not be expected, the evidence from hardness and normalised strength characteristics would support the existence of a miscibility gap in the rhodium-palladium system. The critical temperature of the solid state miscibility gap in the rhodium-palladium binary system is quoted at 830°C and is relatively flat between 30 and 80 atomic per cent rhodium (27). The extent to which the miscibility gap penetrates the ternary system has not been determined but may only be up to approximately 10 atomic per cent platinum, with a critical composition around platinum-40 palladium-50 rhodium, in weight per cent.
High Temperature Properties
Simple binary solid solution platinum-rhodium alloys have a suitable combination of strength and ductility between 1000 and 1500°C to justify their use in many high temperature industrial applications. Many attempts, however, have been made to simultaneously reduce the initial cost and the density of such alloys by the addition of palladium.
Excellent reviews of this subject have been published (11, 25). Although palladium can be added safely to 5 weight per cent rhodium-platinum alloy without loss of high temperature strength, and with a slight improvement in ductility at temperatures up to 1250°C, further substitution for platinum causes a degradation of properties with respect to the parent binary alloy. Figure 3 highlights the effect of palladium substitution for platinum in the pseudo-binary system platinum(palladium)-10 rhodium on both the tensile and 100 hour rupture stress of alloys at temperatures of 1200 and 1400°C.
From mechanical property data reported by Rytvin, Kuz’min and Petrova (8) on a selection of platinum-rhodium-palladium alloys tested at temperatures of 1100 to 1400°C, Figure 4, it is recommended that palladium-rhodium and platinum-palladium-rhodium alloys be used for high temperature applications. The tensile properties of 20, 10, 15 and 7 weight per cent rhodium-platinum binary alloys, however, still exceed those of any other alloy investigated. Rytvin and colleagues report tensile strength values for the platinum-70 palladium-10 rhodium alloy at 1200 and 1400°C to be 22.56 and 5.88MPa, respectively. Reinacher, however, reports respective values of 28.24 and 12.75MPa (10). These differences reflect the variation in pre-test annealing conditions and test sample geometries. Per cent tensile elongations at fracture for a number of ternary alloys have been reported (8, 10), and are shown in Figures 5 and 6. Again there is a striking difference in the values of elongation at fracture for the platinum-70 palladium-10 rhodium alloy. It is reported that the platinum-70 palladium-10 rhodium alloy fails in an intercrystalline manner at 1400°C, but with an elongation at fracture equivalent to that of pure platinum (10).
Perhaps the most important observation made on the high temperature tensile properties of the platinum-palladium-rhodium alloys with palladium contents between 40 and 70 weight per cent, and rhodium contents between 10 and 20 weight per cent, is the low ductilities shown by these alloys between 900 and 1100°C. Values of the per cent reduction in area after tensile test give some indication of this effect and are shown in Figure 7, after (10). An explanation of this behaviour and an interpretation of the data has been given (11).
Creep and Stress Rupture Properties
The stress rupture characteristics of alloys in the platinum-palladium-rhodium system, with rhodium contents up to 20 weight per cent have been thoroughly investigated (7, 10, 16) and previously reviewed (11, 25). There is no doubt that the high temperature performance of platinum-rhodium binary alloys containing up to 20 weight per cent rhodium exceeds that of any ternary alloy examined to date. Ternary alloys which have been investigated fall into two catagories; (i) alloys with a combined rhodium and palladium content not exceeding 20 weight per cent, and, (ii) alloys containing between 40 and 70 weight per cent palladium. A comparison between the 100 hour stress to rupture properties of example materials from each of these categories can be seen in Figure 8, (7, 10). Figures 9 and 10 present 100 hour stress rupture data on selected binary alloys, and pure elements for reference.
At temperatures up to 1250°C, the properties of low-palladium ternary alloys, Figure 8, compare favourably with binary platinum alloys containing up to 5 weight per cent rhodium, Figure 9. A slight improvement in ductility has also been observed.
At temperatures between 1250 and 1500°C, alloys containing less than 5 weight per cent palladium and with rhodium contents below 10 weight per cent show an increase in strain compared to the binary alloys platinum-5 rhodium and platinum-10 rhodium, Figure 11. An increase in the palladium content to 10 weight per cent, however, decreases rupture strain and marginally reduces rupture strength (7). The effect of palladium on the 100 hour rupture stress of a platinum-5 rhodium alloy at temperatures between 900 and 1500°C is shown in Figure 12.
The results of investigations on ternary platinum-palladium-rhodium alloy replacements for the most commonly used platinum-rhodium binary alloy, that is platinum-10 rhodium, have lead to the development of a compromise alloy, namely platinum-3 palladium-7 rhodium. The stress rupture properties of this particular alloy are presented in Figure 13 and, by reference to Figure 9, it is only at operational temperatures of 1500°C that the material shows inferior properties to that of platinum-10 rhodium. An unfortunate consequence of using palladium as an alloying additive to binary platinum-rhodium alloys is the tendency of most materials to fail by intercrystalline cracking at intermediate temperatures between 850 and 1000°C. The lack of ductility in this temperature regime restricts their use to low stress applications.
As far as the high palladium alloys are concerned, results on the stress rupture characteristics of selected materials obtained by Reinacher would suggest that the platinum-40 palladium-10 rhodium alloy has adequate strength properties for industrial applications where the operating temperature does not exceed 1200°C (10, 16). Figures 14 and 15 compare the stress rupture properties of high palladium ternary alloys at 1200 and 1500°C, respectively.
An attempt to correlate the high temperature creep properties of a selection of platinum-palladium-rhodium alloys with palladium and rhodium concentration has been reported (9). Reasonable agreement between experimentally determined values for rupture life and creep rate against calculated values was found. The experimental creep rate data from this work have been reproduced in Figure 16.
The effects of alloying platinum with palladium, rhodium, iridium and ruthenium on the formation of dislocation structures in binary alloys has been investigated by Usikov, Stepanova and Rytvin (17), and highlights the resulting effect on creep rate, Figure 17. Clearly, palladium additions to platinum up to 10 weight per cent have virtually no effect on the stacking fault energy of platinum, with the result that there is little difference between the creep properties of these materials. The effectiveness of other platinum group metals in reducing the stacking fault energy of the platinum solvent and thereby improving high temperature creep properties was found to be ruthenium>iridium>rhodium>palladium. This behaviour is undoubtedly reflected in ternary and higher order alloys and clearly offers the opportunity for developing new alternative heat resistant alloys.
Electrical Resistance and Thermal EMF
The electrical resistance and temperature coefficient of resistance of a complete range of alloys in the platinum-palladium-rhodium system have been measured and reported (6). Iso-resistance plots at 25°C are provided together with a similar graphic presentation for the temperature coefficient of resistance. The highest electrical resistance and lowest temperature coefficient of resistance is shown by an alloy platinum-50 palladium-30 rhodium. Thermal e.m.f. values for alloys containing up to 10 weight per cent rhodium against platinum were determined for the 100 to 1000°C temperature range, and are presented in the form of pseudo-binary sections and a constant thermal e.m.f. contour map. Thermal e.m.f. results are consistent with the view that the platinum-palladium-rhodium ternary system shows complete solid solubility.
The effect of thermal cycling on the creep ductility of platinum-rhodium, platinum-palladium-rhodium and platinum-palladium-rhodium-ruthenium alloys has been studied by Medovoi and Rytvin (13). Sheet samples of alloy were held between ceramic plates and heated by electrical resistance. The thermal cycle was between 1400 and 1350°C, with a full cycle lasting 30 seconds. Time spent at the lowest temperature was 5 seconds. The number of cycles required to achieve a crack length of 1 mm was used as an indication of thermal shock resistance. Ductility was determined from the creep time at 1400°C with an initial stress of 4.9 MPa. The results of this investigation are presented in Figure 18. The binary platinum-rhodium alloys and also the platinum-15 palladium-5 rhodium alloy had the highest thermal shock resistance of all the alloys tested. The basic conclusion from this work was that alloying elements which cause a reduction in grain size and creep ductility of platinum alloys also lower the resistance to thermal shock in the range 1350 to 1400°C.
In 1975, Stepanov, Chernyshova and Shevelev reported on the behaviour of a platinum-15 palladium-5 rhodium alloy and two binary palladium-rhodium alloys during standard welding operations (14). It is clear from this investigation that an alloy containing palladium is more prone to weld crack formation than platinum-rhodium binary alloys. The effects of micro-impurities on weld crack formation were also investigated. The presence of silicon, calcium and aluminium was found to exacerbate the cracking phenomenon in all alloys investigated.
Not only does the presence of metallic impurities affect the properties of platinum group metal alloys but gaseous impurities can significantly alter in-service behaviour. Kalinyuk and Solomentsev have investigated techniques for the determination of gaseous impurities in noble metal alloys (15). Oxygen and hydrogen determinations are presented for platinum and platinum-rhodium, platinum-palladium-rhodium and platinum-palladium-rhodium-iridium alloys. However, no alloy compositions are given as comparative analysis techniques were being evaluated.
Other Ternary and Higher Order Systems
The mechanical properties of two palladium alloys containing rhodium and ruthenium, palladium-1 rhodium-4 ruthenium and also palladium-3 rhodium-2 ruthenium, have been reported by Wise and Eash (4, 5). Ruthenium is shown to be extremely effective in hardening palladium and increases the annealing temperature from 800/900°C to 1000/1100°C. The tensile strength of palladium-1 rhodium-4 ruthenium in the fully annealed state is approximately 90 per cent greater than that of palladium. An increase of 42 per cent is shown for the palladium-3 rhodium-2 ruthenium alloy. Both alloys exhibit reductions in cross-sectional area of greater than 84 per cent in either the annealed or as-received condition. These authors also report a precipitation hardening effect at annealing temperatures of 800°C for the high ruthenium alloy which is similar to that observed in the binary platinum-20 iridium alloy. The internal oxidation of ruthenium may be the cause of such behaviour, resulting in an oxide dispersion strengthened material.
The 1400°C isothermal cross-section of the palladium-rhodium-ruthenium phase diagram has been reported by Raevskaya, Vasekin and Sokolova (3), indicating the existence of a two phase region based upon the binary solid solution of platinum-rhodium and the ternary terminal solid solution of ruthenium. Isothermal cross-sections at 1400°C for the ternary systems iridium-rhodium-ruthenium, platinum-palladium-iridium and palladium-rhodium-iridium are also presented. From microstructural and X-ray analyses it was shown that a complete series of ternary solid solutions were present in the platinum-rhodium-iridium and platinum-rhodium-palladium systems at 1400°C, and that the miscibility gaps in the palladium-iridium-platinum(rhodium) systems were due to decomposition of the solid solution in the palladium-iridium system at 1480°C (27).
Hardness variations are also presented as a function of alloy composition in the palladium-iridium-rhodium and rhodium-ruthenium-palladium systems, and are used to confirm the position of phase boundaries. Some indication of the constitution of the platinum-ruthenium-rhodium-palladium(iridium) quaternary systems has been provided based upon electron probe microanalysis, hardness and lattice cell volume changes of selected experimental alloys.
Similar constitutional work on the ruthenium-platinum-rhodium(palladium)(iridium) ternary systems has been reported by Raevskaya, Vasekin, Konobas and Chemleva (19).
In attempting to define optimum compositions in complex alloyed platinum alloys for use in heat resistant applications, Norvik, Rytvin, Medovoi and Ulybysheva examined statistically a range of alloys in the platinum-palladium-rhodium-gold-iridium system (34). The results of creep rupture tests carried out at 1400°C would initially indicate that a platinum alloy of nominal composition, platinum-10 palladium-10 rhodium, with 0.1 weight per cent of gold and iridium was a suitable candidate for further development. This particular alloy tested at 1400°C under an applied load of 4.9MPa, demonstrated a rupture life capability of 77 hours which is indeed comparable with the properties of platinum-40 palladium-20 rhodium, Figure 15, and platinum-10 rhodium. It was also clearly demonstrated that time-to-failure increased with a decrease in the palladium content and an increase in the rhodium and gold contents. The need for an iridium content of 0.1 per cent was stressed. Further work identified an alloy, platinum-60 palladium-10 rhodium containing 0.1 per cent gold, and excluding iridium, which had a rupture life of 60 hours. If this is indeed the case, then the effect of gold on the stress rupture properties of platinum-palladium-rhodium alloys is quite remarkable, Figure 15, as the rupture life quoted by Reinacher for the platinum-60 palladium-10 rhodium alloy was 1.25h (10).
Considering the importance of the platinum group metals in many high temperature industrial applications, it is surprising that there has been only limited systematic study of the constitutional characteristics and properties of ternary and higher order systems. What is available in the literature has provided important guidelines for those involved in the engineering development of platinum group metal products. Clearly there is a need for further studies of these systems which may result in the development of additional alloys suitable for use in aggressive environments.