Aids in Powder THE ROLE OF THE PLATINUM METALS IN THE ACTIVATED SINTERING OF By C. W. Corti Johnson Matthey Technology Centre

When a metallic powder is subjected to a suffkiently high pressure a certain amount of adhesion takes place between individual particles. If this compact is then sintered the bond is improved by diflusion and inter- granular grain growth. The earliest known platinum objects were fabricated by such a powder metallurgical process, and when European scientists first addressed the problem of platinum bars they also used powder metallurgy to Overcome their inability to melt the metal. Now powder metallurgical methods are widely used for fabricating a variety of materials, and this paper reviews studies made of the sintering of refractory metals when this process is promoted by the addition of a minor amount of a platinum group metal activator.

The manufacture of metals and binder which may be present in sub- alloys in fabricated forms generally commences stantial amounts, for example up to 40 per cent with the melting and casting of ingot material by weight. for subsequent shaping by mechanical tech- In contrast, Activated Sintering is performed niques, such as , rolling and , in the presence of small amounts of metal addi- although in many instances molten metal can be tives, again often transition metals, but in the cast directly to a final shape. However, in the solid state at temperatures below the melting case of the refractory elements, such as point of the additive. Thus, as can be seen from , molybdenum and rhenium, their Table I, Activated Sintering can be accom- very high melting points (in excess of 2o0o0C), plished at lower temperatures than Liquid as well as their resistance to deformation, Phase Sintering, although not necessarily so, generally precludes the melting approach as a depending on the particular metal additive practical route to material and component used. In both cases, however, the temperatures manufacture. This has led to the development employed are substantially lower than would of processes in which consolidation of powder otherwise be required if the refractory powders materials is achieved by sintering at were sintered without additives. Kurtz, for temperatures below their melting points. example, showed in 1946that 99 per cent dense To promote and assist the sintering process, tungsten parts could be achieved by sintering two techniques have been developed which below 140o0C with less than I wt. per cent involve the use of metallic sintering additives. addition of (I),whereas temperatures These are known as Liquid Phase Sintering and above 2800OC are required to achieve a com- Activated Sintering. In Liquid Phase Sintering parable density in untreated tungsten powder. the refractory metal powders are sintered in the presence of one or more metals-generally tran- Activated Sintering sition metals such as or iron-at tem- Since Vacek reported the enhancement of peratures above the melting point of the sintering by additions of small quantities of additive, so that sintering occurs in a molten transition metals to tungsten in 1959 (2),

Platinum Metals Rev., 1986, 30, (4), 184-195 184 Tabla I Typical Sintering Temperatures for Activated Sintering and Liquid Phase Sintering of Refractory Metals

Activated sintering Liquid phase sintering

IMelting Sintering Sintering Refractory point, temperature, temperature, element OC Additive OC Additive OC

3410 Nickel 1100 Nickel 1550 Palladium 1100 Copper 1100 Molybdenum 2610 Nickel 1200 Nickel 1460 PalI ad iu m 1200 - - - ‘ >1350 ‘The tungsten carbide-cobalt eutectic temperature is 1 32OoC making it possible to lower the sintering did not produce any further enhancement; in- temperature substantially, a great deal of work deed, there was a tendency for the sintering rate has been carried out into the activated sintering to decrease from the optimum. They also found of tungsten and other refractory metals, par- that densification occurred in two stages, the ticularly with additions of Group VIII tran- second stage coinciding with the onset of grain sition metals. Much of this work has involved growth in the tungsten (4). the use of platinum group metals which have In an attempt to clarify the mechanism of been shown to be very effective as sintering activated sintering, which they had earlier activators. This paper reviews the published attributed to the activating metal acting as a work on the effect of the platinum group metals carrier phase for the diffusion of tungsten to the on the activated sintering of tungsten and other interparticle “necks”, Hayden and Brophy refractory metals, in particular on the kinetics examined the influence of , and mechanisms of sintering. The properties , platinum and palladium additions on and microstructure of the sintered materials are the kinetics of sintering in the temperature also examined. Finally, the scientific basis for range 850 to IIOOOC(7). the beneficial effect of the platinum group As in their previous work, the platinum metals in the activated sintering of the re- group metal sintering additives were added to fractory metals is examined in terms of current the tungsten powder in the form of aqueous theories and phenomenological models. solutions of salts (chlorides and nitrates) in the requisite amount; this was dried at I~oOCand Tungsten prereduced in hydrogen at 80o0C to form a Much of the early work on the activated metallic coating on the tungsten powder. sintering of tungsten was carried out by Sintering was carried out under hydrogen. Brophy, Hayden and co-workers (3-7). Their For all the platinum group metals examined, initial work focused on the sintering of tungsten Hayden and Brophy found that a minimum powder coated with nickel. They found that, on level of platinum group metal was required to sintering at I IOOOC,the tungsten underwent promote full activation, see Figure I, as had rapid densification to more than 90 per cent been observed in the case of nickel, and that theoretical density. Moreover, they found that larger amounts did not produce any further the amount of nickel required to promote this enhancement. Interestingly, palladium was the accelerated sintering was roughly equivalent to most effective element; this is clearly illustrated a nickel coating thickness of about I atom in Figure 2 which shows the temperature monolayer (3). Nickel coatings thicker than this dependence of shrinkage for each platinum

Platinum Metals Rev., 1986, 30, (4) 185 coating layer. In the case of rhodium there is a transition in the rate controlling process, from the dissolution of tungsten in the rhodium layer at low shrinkages to interface diffusion at large shrinkages. Table I1 summarises these results, the slope, S, being the time dependence of the sintering curves. Also shown in Table I1 are the calculated activation energies for each platinum group metal additive. These lie in the range 86 to I 14 go001 l'o 2.0 kcal/mol, which the authors believed to be

0.002 comparable to the activation energy for tungsten grain boundary self-diffusion. 310 4' AMOUNT OF ADDITIVE,weight per cent The effectiveness of the platinum metals and Fig. 1 During the activated sintering of nickel in promoting enhanced sintering of tungsten the linear shrinkage is dependent tungsten were found to be in the order: upon the activator, and the amount added; Pd > Ni > Rh > Pt > Ru a minimum level beiirequired to promote full aetivation. Data from Hayden and The reason for the platinum group metals being Brophy (7). such effective activators for the sintering of Sintered for 1 hour tungsten was not established in this work, Palladium, 95OOC although it was suggested that it may be linked Ruthenium, 1 100 O C Platinum, 1 100 O C to their relatively high (10 to 20 per cent) Rhodium, llOO°C solubility for tungsten and their low solubility in tungsten. Subsequently Hayden and Brophy extended group metal additive after sintering for I hour. their work on platinum group metal activators Ruthenium was the least effective element. Significantly, the authors found that palladium was better than nickel in promoting densi- fication. For example, the densities of samples sintered at 110o0C for 30 minutes and 16 hours were 93.5 and 99.5 per cent, respectively, in the 0.1. case of palladium in tungsten compared to 92 :005. and 98 per cent, respectively, for nickel in - 4 tungsten. Untreated tungsten would only be 0W presintered at this temperature. 2 0.02- -z Analysis of the sintering kinetics in terms of K the process controlling mechanism in their 5 0.01. carrier phase model of activated sintering- Rhodium ,' ./' 50.005 (lWt.%),' +/ which applies also to liquid phase sintering- / showed that for all the platinum group metals 0.002i examined, the sintering rate was not dependent 1. 800 900 *loo0 1100 upon composition, but was proportional to the TEMPERATURE C cube root of time, except for rhodium in a low Fig. 2 The dependence of linear shrink- shrinkage regime. This time dependence was age upon temperature during the activated interpreted in terms of the diffusion controlled sintering of tungsten. This shows that after sintering for 1 hour, palladium is the most transport of tungsten in the interface between effective activator (7) the tungsten and the platinum group metal

Platinum Metals Rev., 1986, 30, (4) 186 Table I1 Summary of the Effect of Platinum Group Metal Additives on the Sintering of Tungsten (Data from References 7 and 3)

Control Activation energy, Additive Slope ’S’ of process kcal/mol Palladium 0.33 Diffusion 86 Ruthenium 0.39 Diffusion 114 Platinum 0.33 Diffusion 92 Rhodium (a) 0.5 Solution 85 Rhodium (bl 0.33 Diffusion 98 Nickel 0.5 Solution 68

(a) Rhodium : low shrinkage regime Ibl Rhodium : high shrinkage regime to the sintering of tungsten with addi- applicable; rather, they favoured a mechanism tions (8). In contrast to the other platinum in which the surface diffusion of tungsten on metals, they found that iridium actually the activator surface is the controlling step; decreased the rate of densification of tungsten, both are shown schematically in Figure 3. the effect reaching a minimum value at about 2 The influence of a wide range of transition wt. per cent iridium, larger additions having no metal additions, including all the platinum further effect. The measured activation energy group metals, on the sintering of tungsten at of 133 kcal/mol was close to that for volume temperatures between 1000 and 2000OC was self-diffusion of tungsten (I35 kcal/mol). studied by Samsonov and Jakowlev (10). They Further work on the activated sintering of found, in agreement with earlier findings, that tungsten by palladium and nickel additions was additions of Group VIII elements-including carried out by Toth and Lockington, who also the platinum group metals-promoted densi- found that there were optimum concentrations fication of tungsten, with the exception of of both palladium and nickel for maximum osmium which was neutral. Iridium had a small densification during sintering at 10ooOC (9). beneficial effect at the highest temperature Calculations showed these optimum concentra- studied, 2o0o0C, which is not inconsistent with tions to correspond approximately to a mono- the earlier work of Hayden and Brophy (8), layer of the activating element on the tungsten since extrapolation of their Arrhenius plots surface, as also found earlier by Brophy, predicts a transition from a detrimental to a Shepherd and Wulff (3). Once again, palladium beneficial effect at temperatures above about was found to be more effective than nickel, 14moC. The effectiveness of the platinum especially at and below a temperature of 95Ooc. group metals in enhancing sintering was found Toth and Lockington found the time depend- to be in the order: ence of the densification to be 0.5 for both Ru < Rh < Pd palladium and nickel, in contrast to the value of and 0s < Ir < Pt 0.33 for palladium found by Brophy (7). The with the upper row of elements being superior apparent activation energies were lower, 62.5 to the lower row. This is shown in Table 111, kcal/mol compared to 86 kcal/mol for which also gives the measured values of com- palladium and 50.6 kcal/mol as against 68 pressive strength, hardness and grain size. On kcal/mol for nickel. Microprobe analysis of the this basis, nickel appears to be slightly more fracture surfaces of sintered specimens showed effective as an activator than palladium, in con- segregation of the activating elements on grain trast to the earlier work, but this is based on boundary surfaces. The authors concluded that results obtained at higher sintering tem- Brophy’s model for activated sintering was not peratures than those of the earlier studies.

Platinum Metals Rev., 1986, 30, (4) 187 Fig. 3 These two models show different representations of the sintering of tungsten particles which have been coated with a metallic activator. (a) The model of Brophy, Hayden and WN(3) has tungsten diffusing through the carrier (activator) phase, away from the line joining the centres of adjacent particles, to be redeposited elsewhere on TUNGSTEN the particles as indicated by the arrows. (b) The model of Toth and Lockington (9),where dissolution of tungsten at the activator-tungsten interface is followed by volume diffusion outwards through the activator layer and subsequent sur- face diffusion, this being the rate con- trolling step. Diffusion through the activator layer to the contact point bet- ween adjacent particles results in the for- mation of sintering “necks”

These results show that the stronger activators Figure 4. The arrows indicate an increasing also enhance the associated grain growth in the degree of activation. They interpreted these final stage of sintering. The higher densities results in terms of the electron structure of the (lower porosity) achieved are also reflected in activators and tungsten; an increase of the higher values of compressive strength and stable d-bonds in the system lowers the free hardness. energy, activating the sintering process in Samsonov and Jakowlev summarised their which diffusion is accelerated by the activators findings in terms of the position of the for which tungsten acts as an electron donor. activating element in the Periodic Table, More recently, German and his co-workers

Tabla 111 Dependence of the Properties of Tungsten on the Activating Element at the Optimum Concentration and Sintering Temperature (10)

Activator, Sintering Compressive Tungsten weight temperature, Density, strength, Hardness, grain per cent OC g/cm3 kg/mm* kg/mm2 size, pm w 2000 16.1 80 181 5-7 Fe (0.5- 1 .O) 1600 17.4- 17.95 92-96 310-390 12-15 CO (0.3-0.4) 1600 17.4- 17.95 83-87 277-282 20-25 Ni (0.2-0.4) 1 400- 1600 18.1- 18.4 91 -95 280-306 100- 1 20 Ru (1.0) 1600- 1800 17.4 77 309 15-20 Rh (0.5) 1600 17.8 71 290 10-1 5 Pd (0.3-0.4) 1400- 1600 18.1 -18.35 83-90 290-300 20-25 0s (1.0) 2000 16.1 80 181 5-7 Ir (1.0) 2000 16.7 98 238 5.5-7 Pt (1.0) 1800-2000 17.6- 17.9 88-93 305-330 20-30

Platinum Metals Rev., 1986, 30, (4) 188 have investigated the activated sintering of (I I - PALLAOIUM,wetght per cent tungsten in more detail 13).German and 0.01 0.1 Ham (I I) confirmed that palladium is the best metallic activator for the sintering of tungsten NICKEL.Welght per cent 0.01 0.1 in the range 1100 to 140o0C, as shown in 1 Figure 5. This shows that, for both palladium and nickel, enhancement of sintering starts at approximately I monolayer thickness of addi- tive and peaks at a thickness of 4 monolayers. Sintering in a moist hydrogen atmosphere was found to be detrimental to palladium activation

TI y

EOUIVALENT MONOLAYER THICKNESS Fig. 5 The effect of the thicknees of the palladium or Rickel activators upon the linear shrinkage of tungsten powder sintered at 1200 and 130OOC in dry hydrogen is shown, derGerman and Ham (11) Tungsten+palladium, 130OOC A Tungsten + nickel, 1300OC v Tungsten+palladium, 120OOC 0 Tungsten+nickel, 1200OC

. for one tungsten powder, but beneficial for a second. The apparent activation energy is lowered on sintering in a moist atmosphere. German and Munir (12)extended this work to other Group VIII elements including platinum, and confirmed that enhanced sinter- ing commenced at about I monolayer thickness and peaked at 4 monolayers. They found the effectiveness of the activator to be in the order: Pd>Ni>Pt = Co>Fe>Cu Below 130o0C, iron was more effective than platinum and cobalt. In the case of palladium and nickel, where sintering progressed to the second stage, extensive tungsten grain growth Fig. 4 Trends in the activated sintering of tungsten are related to the position of the was observed. The onset of grain growth was activating element in the Periodic Table, as associated with a decline in the shrinkage rate. proposed by Samsonov and Jakowlev (10). The authors found a time dependence of the l%e arrows indicate increasing degrees of sctivation shrinkage for all activators, similar to that found by Toth and Lockington, favouring

Platinum Metals Rev., 1986, 30, (4) 189 volume diffusion of tungsten through the much higher than in prerecrystallised tungsten, activator layer as the rate controlling which is attributed to the high diffusivity paths mechanism. The addition of 0.4 weight per through an intergranular phase formed by the cent palladium was found to increase the activator which segregates to the grain apparent grain boundary diffusion rate by boundaries. Auger electron spectroscopy about 6 orders of magnitude, corresponding to revealed this layer to be about 2nm thick for a decreased activation energy. The measured both palladium and nickel. The measured dif- activation energies decreased in the order of fusivities of the activators were in the order: increasing activator effectiveness, palladium Pd > Ni > Pt > CO having the lowest value. This was related to the Studies by Gessinger and Buxbaum on elec- electron structure modifications as postulated tron emission from thoriated tungsten cathodes by Samsonov and Jakowlev (IO), that is the has shown that platinum can also activate transition metals with unfilled d-shells are the enhanced diffusion of thorium to the surface optimal activators for tungsten. Based on this along grain boundaries, enabling the tempera- concept, the authors suggest that palladium and ture limit for electron emission to be extended nickel are optimum activators for all refractory metal powders. Later work by Li and German examined the 80, 1 properties of palladium- and nickel-activated Pal lad i urn tungsten sintered with optimum activator con- tent in the temperature range 1200 to 16ooOC (13). Hardness levels were in the order Untreated Pd > Ni > Co > Fe at lower temperatures, as can be seen in Figure 6, but were closer together at the higher temperatures. In the case of transverse rupture strength, nickel-activated tungsten was stronger than palladium-activated tungsten, the strength decreasing with in- I I creasing sintering temperature above 140ooC due to rapid grain coarsening. For the 0.43 +I weight per cent palladium-activated material, 2 the grain size increased from 4.5pm at 1200~C K400- cv) to 18.opm at 140oOC and to 28.5pm at 1600OC. W Recent work on activated secondary re- 3 300- crystallisation of heavily-drawn doped tungsten U wire has provided additional evidence for the W U influence of the activators during sintering (14). 9 200- In this work the tungsten wire was coated with z E palladium, platinum or nickel prior to anneal- W ing and the rate of secondary recrystallisation 0 100. UW measured. The highest rate of recrystallisation P was found in the presence of palladium, fol- 1200 1400 1600 lowed by nickel and then platinum, grain SlNTERlNG lEMPERATURE,'C growth being induced at temperatures several Fig. 6 The hardness and the strength of hundred degrees lower than uncoated tungsten tungsten after activated sintering with wire. The process was controlled by the pene- various activators is shown here as a func- tration of the activating elements into the wire. tion of sintering temperature, from Li and German (13) The diffusivities of these were found to be

Platinum Metals Rev., 1986, 30, (4) 190 from 1950 to 215oK and the maximum grain boundary heterodiffusion model emission current to be increased from 3 to 7.5 developed earlier for tungsten (21). A/cm2 (IS). This work demonstrates that Later studies by German and Labombard platinum group metals not only enhance the (20) on palladium, nickel and platinum diffusion of tungsten, but can also enhance the additions to two different molybdenum diffusion of other elements in the tungsten powders of the same particle size sintered at low grain boundaries. temperatures (1050 to I I 50°C) confirmed the earlier findings, namely that palladium is the Molybdenum and Other Refractory best activator, followed by nickel and platinum, Metals and that sintering behaviour conformed to the As with tungsten, several investigators have heterodiffusion model. shown that both palladium and nickel can German and Munir also studied the activated enhance the sintering of molybdenum, for ex- sintering of hafnium (22) and tantalum (23) ample see References 16 to 18. Further, more with transition metals as part of their broader detailed work on the activated sintering of investigation into the mechanisms, particularly molybdenum by platinum group metal addi- the d-electron exchange model proposed by tions has been carried out by German and his Samsonov (10). In the case of hafnium, the ac- co-workers in the U.S.A. (19, 20). Their work tivators were added to a thickness equivalent of on the heterodiffusion modelling of tungsten 4 monolayers. Isothermal sintering experiments was extended to molybdenum where the effect in the range 1050 to 145oOC showed densi- of I 3 transition metal additions including fication after I hour in the following order of rhodium, palladium, iridium and platinum was enhancement: examined in the temperature range 1000 to Ni > Pd > Co > F’t 135oOC (19). Again, they found that activation Cobalt and platinum were only beneficial at of sintering commenced at activator concentra- temperatures above about 130oOC. Unusually, tions equivalent to I monolayer thickness and a non-Arrhenius temperature dependence was reached the maximum effect at about 10 found for the activated sintering, and this was monolayers’ thickness, although this plateau confirmed by constant heating rate experiments shifted to greater thicknesses with increasing which showed sharp peaks for some activators sintering temperature. They confirmed that at varying temperatures. These peaks occurred palladium was the best activator for at about 1375OC for palladium and about molybdenum, with the degree of effectiveness 1240OC for nickel, for example, and are indica- being in the order: tive of an optimum sintering temperature for Pd > Ni > Rh > Co > Pt > Au > Fe maximum densification enhancement. It was As in the case of tungsten, iridium was concluded that the activated sintering of haf- detrimental to the sintering of molybdenum. nium does correlate with the electron structure The activation energy for sintering decreased model, although the activators only impart a with increasing effectiveness of the activator, limited benefit. that for palladium-activation being 280 k J/mol In their study of the activated sintering of (66.9 kcal/mol) compared to 405 kJ/mol (96.8 tantalum in the range 1250 to 170ooC, German kcal/mol) for untreated molybdenum. This and Munir found only slight enhancement with decrease in activation energy for palladium, platinum, palladium and rhodium (23). The nickel and platinum was observed to be concen- poor enhancement of palladium was particular- tration dependent, a rapid decrease occurring at ly surprising, but examination of fractured sur- about I monolayer thickness and reaching a faces suggested that palladium enhanced only minimum value at the optimum concentration surface diffusion, not bulk diffusion. of about 10 monolayers. The authors concluded The use of palladium and nickel as activators that the sintering process was in accord with the in the sintering of chromium has been studied

Platinum Metals Rev., 1986, 30, (4) 191 by substantial grain growth, grain sizes of 10to I 5pm being observed compared to I to 2pm for untreated rhenium. Sintering at 180o0C pro- duced densities of 92 per cent in 0.2 wt. per cent palladium-activated rhenium and only 81 per cent for untreated rhenium. In their study on the sintering of rhenium, German and Munir found that at 10ooOC only platinum enhanced sintering, while elements such as palladium, nickel, iron and cobalt inhibited densification (26). The enhancement effect of platinum commenced at a thickness of about I monolayer, reaching a peak at about 2 monolayers. At 140ooC, both platinum and palladium enhanced sintering, platinum Fig. 7 This geometric model of the heterodiffueion controlled activated sinter- becoming effective at about I monolayer, rising ing process shows the activator layer wet- to a plateau of maximum effectiveness at about ting the interparticle grain boundary, after 4 monolayers. Palladium acts less rapidly, German and Munir (29) reaching that of platinum at about 10 mono- layers thickness, which suggests that palladium would be better than platinum at higher con- at the Tokohu University in Japan (24). It was centrations. The results of this study were found that palladium enhanced sintering con- considered to correlate with the electron struc- siderably, the degree of enhancement reaching ture model reasonably well. a plateau at about 0.8 wt. per cent palladium The study of palladium additions to co- over the range 1050to 120oOC. Above 120ooC, reduced tungsten-rhenium powders by on sintering for I hour, the extent of enhance- Shnaiderman and Skorokhod again illustrates ment became suppressed; this was mainly an ef- the beneficial effect of palladium in activated fect of the higher density levels achieved at the sintering of refractory metals, although in this higher temperatures and a reflection of the instance palladium-rich interlayers are retardation due to grain growth in the second formed at the grain boundaries (28). stage. In contrast, nickel had little activating effect. The relative behaviour of palladium and Models of Activated Sintering nickel was interpreted by the authors in terms As we have seen, the activation of sintering of the mutual solubility criterion suggested refractory metal powders by transition metal earlier by Hayden and Brophy (7). Palladium- elements has been interpreted in terms of chromium fulfils this requirement whereas several models, which are generally qualitative nickel-chromium does not. in nature. The results of the many studies on The activated sintering of rhenium and several refractory metals have shown a reason- tungsten-rhenium mixtures has been studied by ably consistent pattern in that the most bene- several investigators (25-28). Dushina and ficial activators are palladium, nickel and Nevskaya found that on sintering rhenium for platinum, generally in that order. Since these 2 hours in the range 1300 to 20oooC, substan- three elements sit in the same column of the tial enhancement of sintering occurred with Periodic Table, it is reasonable to assume that palladium contents of 0.I to 0.5 weight per cent their role is related to their electronic structure (25). Maximum enhancement was found in the and its ability to promote rapid diffusion of the range 0.2 to 0.4 wt. per cent palladium. This refractory element. The time, temperature and enhancement was observed to be accompanied activator concentration dependencies are also

Platinum Metals Rev., 1986, 30, (4) 192 MODEL ----- EXPERIMENTAL CONSENSUS - Fig. 8 The position of the activator in the Periodic Table affects the densi- fication of the tungsten. There is good agreement between the predictions of the heterodiffusion model (29) and experimental consensus

similar for all the refractory metals studied, rise to the high solubility in the activating which suggests that there is a common basis for element. a generalised model (29). A more recent proposal by German and The initial model postulated by Hayden and Munir takes this model further (21)and applies Brophy was based on a solution-precipitation the Engel-Brewer theory (30) to the prediction approach in which the relative solubilities of the of the activation energies for the diffusion of activator in the refractory element and the refractory metals through the activator layer. In refractory metal in the activating element this quantitative model the activator has a role should be low and high, respectively (7). In this in providing enhanced grain boundary dif- model, illustrated schematically in Figure 3(a), fusion of the refractory metal. This is shown the refractory element diffuses away from the schematically in Figure 7, with the activator interparticle boundary and is redeposited layer wetting the interparticle grain boundary. elsewhere on the particle surface-as indicated This is taken from Reference 29, where a more by the arrows in the Figure. Experimentally the detailed description of these models is given. rate controlling step is found to be either refrac- The relative solubility criterion is a prerequisite tory metal diffusion at the interface with the ac- for enhanced diffusion of the refractory metal. tivator layer, or refractory metal solution in the Enhanced mass transport, and hence densi- activator layer. fication, results from the lowering of the acti- Later, Samsonov and Jakowlev proposed that vation energy for the refractory metal in the activated sintering was a consequence of the activator. electronic structure stabilisation of the refrac- German and Munir have shown (29) that tory metal caused by the additive metal (10). their calculated values of activation energy for They based this approach on the argument that diffusion of molybdenum agree well with a metallic system containing partially filled experimentally determined values (19). These d-subshells becomes more stable as the number calculations indicate that palladium, nickel and of d5 and d*O electron configurations increase. platinum are the best activators, as shown The refractory element acts as an electron experimentally for several refractory metals. donor, and this ease of electron transfer gives The predicted shrinkages for molybdenum also

Platinum Metals Rev.,1986, 30, (4) 193 agree well with experiment, as shown in Table IV (29). Figure 8 shows the good agreement Table IV between their predictions and the experimental Comparison of Predicted and consensus for tungsten in terms of the position Measured Shrinkage for the of the activator in the Periodic Table. This Activated Sintering of Molybdenum Powder (29) clearly demonstrates the superiority of pal- 2.2pm size: sintered 1 hour at 125OOC ladium as an activator, with rhodium and in hydrogen platinum also in significant positions. Shrinkage, per cent More recently, Miodownik has proposed a I quantitative figure of merit for assessing the Activator 1 Predicted I Measured potential of additive elements as activators (31). Untreated Mo 1.6 2.1 This parameter, 4, has been derived by com- Pd 8.9 8.2 bining the relevant heats of solution, surface Ni 6.9 7.2 Pt 4.7 3.6 energies and the energy of vacancy formation in co 4.6 4.3 the activator, and is based on the underlying Fe 3.5 2.2 thermodynamic parameters that are responsible Cr ~ 1.1 1.1 for the phase equilibria; the solubility criterion of the earlier models is an aspect of the latter: 4 = AH, + AH2 + AH3 presence of activators would be expected to where AH,, AH, and AH, are thermodynamic promote diffusional creep of the Coble type. functions related to solubility, segregation and As stated above, the use of platinum group diffusion, respectively. Using calculated values activators can promote enhanced electrical pro- of for the sintering of tungsten, Miodownik’s perties: Gessinger and Buxbaum have utilised predictions are correct for 12out of 14activator the increased grain boundary diffusivity in elements shown in Figure 8, the only dis- platinum-activated thoriated-tungsten emitters crepancies being manganese and gold. Once to improve electron emission (I5). again palladium is predicted to be the most effective activator. Summary This paper has reviewed the numerous Properties of Sintered Materials studies of the activated sintering of refractory While there has been a considerable number metals by transition metal additions and has of studies into the phenomenon of activated shown that the platinum group metals, and par- sintering, relatively few studies have measured ticularly palladium and platinum, even at the properties of the sintered materials. amounts of less than I per cent by weight, are Fracture is generally intergranular in nature, very effective in promoting densification at suggesting that the activator-rich grain temperatures several hundred degrees lower boundaries are paths of easy fracture. Strength than would otherwise be required. The models and hardness are very density dependent and developed to describe the phenomenon have the effectiveness of the activator on densi- been examined and those of German and Munir fication clearly plays a major role. Thus, both (29) and Miodownik (31) have been shown to type and concentration of the activator predict the order of effectiveness remarkably influence sintered strength, as shown in Table well. Significantly, palladium is predicted to be I11 and Figure 6. Strengths as high as 1050MPa the best activator element for several refractory have been shown for palladium-activated metals including tungsten and molybdenum, in tungsten (10). accord with experimental findings. The No data have been presented for high enhanced densification that results from the use temperature creep properties, but the enhance- of platinum group metal activators leads to ment of grain boundary diffusion in the improved strength properties.

Platinum Metals Rev., 1986, 30, (4) 194 References I J. Kurtz, Proc. Second Annu. Spring Meeting, I7 0. Neshich, V. V. Panichkina and V. V. Metal Powder Assoc., 1946, 40 Skorokhod, International Team for Studying 2 J. Vacek, Planseeber. Pulvermetall., 1959, 7, 6 Sintering, ITS 27, 1972 I8 Skorokhod, M. solonin, L. 1. Cherny- 3 J. H. Brophy, L. A. Shepherd and J. Wulff, v. v. s. “powder all^^^>,, ed. W. hszbski, NME- shev, L. L. Kolomiets and L. I. Schnaiderman, MPI, Interscience, New York, 1961,p. 113 Soviet Powder Metall. Ceram., 1976, 15, 435 R. M. German and Z. A. Munu, Less-Common 4 J. H. Brophy, H. w. Hayden and J. wulff, 19 J. Trans. AIME, 1962, 224, 797 Met., 1978,58, 61 R- M. German and A. In[.9. 5 H. W. Hayden, S.B. Thesis, Massachussetts 2o c. Institute of Technology, Metallurgy hpt., 1960 PowderMetaN. Powder Technol-, 1982, 1% (2), 147 6 J. H. Brophy, H. W. Hayden and J. wulff, 21 R. M. Gem,“Sintering-New Developments”, Trans. AIME, 1961,221, 1225 4th Int. Conf. on Sintering, ed. M. M. Rustic, Dubrovnik, Sept. 1977,Elsevier, 1979,p. 257 7 H. W. Hayden and J. H. Brophy, 3.Electrochem. 22 R. M. German and A. Munir, 3. Less-Common Sot., 1963,11% (7), 805 2. 8 H. W. Hayden and J. H. Brophy, J. Less- Met.’ 1976’46’ 333 Common Met., 1964, 6, 214 23 R. M. German and Z. A. Munir, Powder Metall., 145 9 I. J. Toth and N. A. Lockington,J. Less-Common 1977y20’ (3)’ 24 R. Watanabe, Toguchi &d Y. Masuda, Sci. Met., 1967,12, 353 K. I0 G. W. Samsonov and W. I. Jakowlev, Z. 1983’Is’ (’)’ 73 Metallkd., 1971,62, (S), 621 25 0. V. Dushina and L. V. Nevskaya, Soviet Powder Metall. Ceram., 1969,8, 642 I1 R. M. German and V. Ham, Int. J. Pmder Z. Metall. Powder Technol., 1976, 12, (2), 115 26 R. M. German and A. Munir, J. Less-Common I2 R. M. German and Z. A. Munir, Metall. Trans. Met‘’ 1977’53y 14’ A, 27 V. V. Panichkina, L. I. Shnaiderman and V. V. 1976,7A, 1873 Skorokhod, Dokl. Akad. Nauk. Ukr. SSR,’ C. Li and R. M. German, Metall. Trans. A, 1983, 1975, 13 (9,469 14,2031 A, 28 L. I. Shnaiderman and V. V. Skorokhod, Soviet I4 R’ Warren and E’ Henig3 to be Powder Metall. Ceram., presentedL. at the 7th Int. Risd Symposium, 1980, 19, (I), 27 Roskilde, Denmark, Sept. 1986 29 R. M. German and 2. A. Munir, Rev. Powder Metall. Phys. Ceram., 1982,2, (I), 9 I5 G. H. Gessinger and Ch. Buxbaum, “Sintering and Catalysis”, ed. G. C. Kucqnski, Plenum, 30 L. Brewer, “High Strength Materials”, ed. V. F. New York, 1975, p. 295 Zackay, J. Wiley, New York, 1965, p. 12 16 V. V. Panichkina, V. V. Skorokhod and A. F. 31 A. p. Miodownik, “Shtering: Theory and Prac- Khrienko, Soviet Powder Metall. Ceram., 1967,6, tice”, Hamwe, October 1984; Metall., 558 1985,28, (31, 152 Oxidation Behaviour of Some Platinum Alloys

For a limited number of specialised applica- change, but no changes were perceived on the tions, such as for jewellery, the aesthetic ap- four platinum alloys, which contained 5 weight pearance of a material is a crucial factor. Clearly per cent cobalt, 10 and 15 palladium, and 7 the appearance must be pleasing at the time of palladium plus 3 cobalt. purchase, and for precious metal items it is Auger electron spectroscopy detected surface equally important that they should not lose segregation of alloying elements on all samples, their appeal with use or the passage of time. except for the alloy containing cobalt, an ele- Recently the results of a study sponsored by ment for which the technique is insensitive. In Rustenburg Platinum Mines Limited into the general, the platinum alloys showed no signifi- oxidation behaviour of a number of commer- cant changes as a result of the oxidation treat- cially available platinum-rich and 18 carat gold ment, although minimal oxidation-enhanced alloys has been reported (A. Wells and I. Le R. copper enrichment was observed on the surface Strydom, J. Muter. Sci., Lett., 1986, 5, (7), of the platinum-15 palladium alloy, in which a 743-746). The reactivity of the alloys was small amount of copper is incorporated. assessed by examining them in both the as- From this study it was concluded that the received condition and after heating at I~oOC platinum alloys examined were significantly for 24 hours under a flow of oxygen. less environmentally reactive than the two gold After this treatment it was observed that both alloys, under oxidising conditions at near am- the gold alloys had undergone a colour bient temperatures.

Platinum Metals Rev., 1986, 30, (4) 195