For this experiment, chloride and the chloride–sulfate solutions of industrial copper and nickel anode sludges were used. Their contents are listed in Tables I and II.

2.1.1 Treatment of Copper Anode Sludge Solution

2.1.2 Treatment of Nickel Anode Sludge Solution

Section 2.1 demonstrated the capability of extracting and concentrating pgms from the solutions of their lean bulk concentrates, for example copper and nickel anode sludges. At the same time, the capability of using hydrocarbonylation to treat the rich bulk concentrates of pgms is also of great technological interest. These concentrates are represented by, in particular, three types (pgm concentrates predominantly containing: (a) platinum and palladium, M-1; (b) rhodium, ruthenium and silver, M-2; (c) iridium, M-3; and (d) speiss alloy). At present, the refinement of such concentrates is conducted under separate processes (1).

To conduct an experimental treatment through hydrocarbonylation, a total chloride solution obtained by the hydrochloration of concentrates M-1, M-2 and M-3 at the ratio of 4:2:1 was used. During the hydrochloration process of the concentrate mixture, silver was removed as AgCl precipitate. Thereafter, from the obtained solution, gold was extracted by treating it with Fe₂(SO₄)₃ solution. Table VII describes the content of the obtained solution. Within the experiment, carbon monoxide was bubbled through 20 ml of the original solution (see Table VII) under atmospheric pressure and temperature of 70°C while stirring vigorously for 3.33 h.

The solution was observed to acquire black colour as a result of the generation of finely dispersed powders of palladium and platinum, which were easily filtered out with a vacuum filter. The filtrate was red and brown. The obtained precipitate was washed with 2 M HCl, dried and weighed. 0.5515 g of black-coloured powder was generated. The precipitate possessed large specific surface area and high absorbing capacity, thereby required thorough washing.

The analysis of the powder and filtrate for the content of palladium, platinum, gold, rhodium, ruthenium and iridium showed almost full precipitation of palladium, platinum and gold (none of the elements were found in the filtrate), no ruthenium and iridium were found in the powder, and the extraction rate of rhodium from the solution amounted to approximately 1%.

It is worth noting that platinum from the solution is not recovered as metal, but as its oligomeric dicarbonyl [Pt(CO)₂]_n, where n is divisible by three. Palladium is extracted in its native form, together with its amorphous carbon phase (4).

In another experiment, 50 ml of the original solution (see Table VII) was treated with carbon monoxide under atmospheric pressure and temperature of 95°C for 4.5 h while stirring vigorously. Finely dispersed black powder was produced. After filtering, washing with 2 M HCl and air drying, 1.3501 g of a black powder were recovered, which was analysed for the content of noble metals (Table VIII).

The filtrate, which was yellow and green, was extracted twice with 10 ml of isoamyl alcohol. The organic phase coloured in yellowish was boiled dry without calcination. The boiled organic phase was analysed for the content of noble metals (Table VIII).

The results of the laboratory treatment of the mixture of concentrates M-1, M-2 and M-3 show that the hydrocarbonylation process can be successfully used at the primary stage of refinement, and this allows to separate rhodium, ruthenium and iridium from other noble metals and makes the subsequent extraction of platinum and palladium, as well as rhodium, ruthenium and iridium easier.

To confirm the possibility of extracting rhodium, ruthenium and iridium using the traditional technology of the nitration of their carbonyl chloride solutions, which are left after extracting the concentrate of platinum, palladium, gold, tellurium and selenium by hydrocarbonylation, a process solution of speiss with the following content (g l^–1) was used: platinum, 20.02; palladium, 53.00; rhodium, 6.50; ruthenium, 0.50; iridium, 2.70; gold, 0.82; tellurium, 3.00; selenium, 2.30; copper, 16.40; lead, 3.70; bismuth, 3.85; nickel, 2.90; iron, 1.55; HCl, 100.00. Carbon monoxide was bubbled through 500 ml of the specified solution at 60°C under atmospheric pressure for 4 h while stirring vigorously. This generated the precipitate of Σ platinum, palladium, gold, selenium, tellurium, which was washed with 2 M HCl after the separation. The precipitate and filtrate were analysed for the content of original components. It was concluded that gold, palladium, selenium and tellurium precipitated fully whereas platinum precipitated by 98%. The solution after the precipitation of Σ platinum, palladium, gold, selenium, tellurium was nitrated with the precipitation of rhodium in the form of its ammonium–sodium hexanitrate: (NH₄)₂NaRh(NO₂)₆.

Thus, the compatibility of hydrocarbonylation and the subsequent processes of the traditional saline extraction technology for rhodium, ruthenium, iridium (1) has been shown.

Technologies for the extraction of pgms from their chlorocomplexes utilise sorption on ion exchange resins (1). Thus, exploring the opportunity of using sorption for the solutions of carbon chloride anionic complexes Rh(I)–Rh(CO)₂Cl₂⁻, Ru(II)–Ru(CO)₂Cl₄⁻² and Ir(I)–Ir(CO)₂Cl₂⁻ to extract rhodium, ruthenium, iridium was appropriate.

To this extent, we used a solution of chlorocomplexes RhCl₆⁻³, RuCl₆⁻² and IrCl₆⁻² containing (mg l^–1): rhodium, 260.3; ruthenium, 83.0; iridium, 63.8 and HCl, 2 mol l^–1, and a solution of carbonyl chloride anionic complexes rhodium(I), ruthenium(II) and iridium(I), obtained by treating a similar solution of chlorocomplexes RhCl₆⁻³, RuCl₆⁻² and IrCl₆⁻² with carbon monoxide at 80°C for 2 h, which led the solution to change its colour from red and brown to yellowish, which is a distinct feature of carbonyl chloride anions rhodium(I), ruthenium(II) and iridium(I).

Both solutions were engaged with gel anion exchange resin based on the copolymer of styrene and divinylbenzene with benzene–pyridinium functional groups (ammonium molybdophosphate (AMP)) while stirring for 2 h. The solutions were then analysed for the content of rhodium, ruthenium and iridium. Results for the observed sorption rate are given in Table IX. The results given in Table IX show that the processes of hydrocarbonylation and sorption are sufficiently technologically compatible.

Given the specific properties of carbonyl chloride complexes rhodium(I), ruthenium(II), iridium(I) (2), it was of interest to obtain the experimental results of their reaction with hydrogen under atmospheric pressure, as it is known that chlorocomplex solutions rhodium(III), ruthenium(IV), iridium(IV) only react with hydrogen at high pressure (5). For this experiment, a standardised test solution of chlorocomplexes rhodium(III), ruthenium(IV), iridium(IV) in 2 M HCl with the following metal concentration (mg l^–1): 168.0; 119.0; 70.0, respectively, was treated with carbon monoxide under atmospheric pressure and temperature of 95°C for 2 h. The solution was observed to quickly lose its original red and brown colour and turned a yellowish colour, which is a characteristic of carbonyl chloride complexes rhodium(I), ruthenium(II), iridium(I). Thereafter, at the same conditions, hydrogen was fed to the reactor instead of carbon monoxide. This provoked the solution to gain black colouration and precipitate. The hydrogen treatment lasted 6 h. The obtained precipitate was filtered out and dissolved in the mixture of HCl and H₂O₂. The atomic absorption analysis of the precipitate and filtrate has shown that the resulting black precipitate comprised rhodium without ruthenium and iridium content, and that the filtrate contained almost no rhodium. The filtrate was neutralised to pH 5 and treated with hydrogen under the same conditions. The solution gained black colour and released a black precipitate. The hydrogen treatment lasted 6 h. The black precipitate was filtered out and dissolved in the mixture of HCl and H₂O₂. The atomic absorption analysis of this solution showed that it contained ruthenium only, whereas the analysis of the solution after filtering out black ruthenium showed that there was almost no ruthenium in it. The yellowish filtrate was treated with chlorine gas, which produced intense red colouration characteristic for chlorocomplex iridium(IV). NH₄Cl in the form of saturated aqueous solution was added to this solution. This resulted in the black precipitation, characteristic for (NH₄)₂IrCl₆. The above-precipitate solution analysis showed trace amounts of iridium.

The described experiment has shaped a foundation for conducting a series of experiments aimed to optimise pgm reduction processes, which allowed to formulate the method for selective extraction of rhodium, ruthenium and iridium from rhodium(I), ruthenium(II) and iridium(I) carbonyl chloride solutions by treating them with hydrogen under atmospheric pressure (6).

Reactions (vi)–(xiii) have some induction period (T_ind.), i.e. time from the beginning of the treatment of the pgm chloride solution with carbon monoxide until black pgm precipitation or change in original colour. The induction period of the hydrocarbonylation is impacted by temperature, stirring speed, partial pressure of carbon monoxide and H⁺ and Cl⁻ ion concentration (2).

Studies were conducted to identify the impact of the carbon monoxide content in the reaction gas (P_CO) on the induction period of a hydrocarbonyl process. As an example, we provide the results of the studies on the impact of P_CO on the induction period of Reaction (vi) (Table X). The original H₂PdCl₄ solution contents were as follows: [Pd(II)], 167 mg l^–1; HCl, 2 mol l^–1; solution volume, 75 ml; CO + N₂ mixture bubbling rate, 100 ml min^–1; stirring speed, 1000 rpm; temperature, 50°C. The length of the induction period of Reaction (vi) was calculated based on the automatic logging of changes in the light transmission value of the solution.

The experimental data provided in Table X indicate that air-producer gas can be used for conducting hydrocarbonyl processes in technological operations, as the significant decrease in the hydrocarbonylation rate is observed when carbon monoxide content in the CO + N₂ mixture does not exceed 20% whereas air-producer gas contains 25% of carbon monoxide.

Platinum-rhodium thermocouples are used widely as temperature sensors (1) and are often used in industry where the accuracy and stability requirements are greater than those that can be provided by less expensive base metal thermocouples. Examples of such processes are iron and steel manufacturing (2), quartz glass manufacturing and aerospace heat treatment and casting. Platinum and rhodium are also important industrial catalysts, for example for the oxidation of ammonia to nitric oxide (3, 4).

At temperatures above about 1200°C, it is known that platinum-rhodium thermocouple wires form volatile oxides where the solid platinum or rhodium oxide films formed on the surface of the wires at lower temperatures evaporate (5). This causes the wires to lose mass (3, 6–8), which is the dominant cause of progressive instability and thermoelectric-inhomogeneity of the platinum-rhodium alloyed thermocouples. This instability arises because the vapour pressures and oxidation rates of platinum and rhodium are different. The overall behaviour also differs for each alloy (2, 6, 9). The departure of platinum and rhodium oxides from the wire at different rates causes a change in the composition of the thermoelements (4, 6) and hence calibration drift of the thermocouple.

An investigation conducted by Rubel et al. (4) studied the contribution of the volatile oxides to the mass loss of Pt-5%Rh and Pt-10%Rh alloy gauzes. The gauzes were exposed to oxygen over periods of 70 h to 300 h at 890°C and 1100°C. By using X-ray photoelectron spectroscopy (XPS), the type of volatile oxides produced were identified as platinum and rhodium oxides PtO₂ and RhO₂ respectively. It was also found that the rate of mass loss of rhodium oxide was greater with increasing rhodium content of the platinum-rhodium alloy.

Although the oxidation rates of platinum-rhodium alloys have not been investigated much in recent years, the effect of oxidation of the platinum group metals at high temperatures has been widely studied. Phillips, in particular, studied the mass loss of the platinum group metals osmium, palladium, ruthenium, platinum, iridium and rhodium, and evaluated the time dependence of the mass loss by considering metal powders placed in 1’’ billets over the temperature range from 600°C to 1400°C (10). He concluded that all the platinum group metals lost mass when oxidised in air and that the rate of mass loss was greater at higher temperatures. Notably, the rate of mass loss increased with temperature, and for a given temperature the mass loss was linear with elapsed time, provided the volatile oxides are formed from the corresponding metal.

Another key experiment to determine the mass loss of platinum metals was conducted by Crookes (11). The two temperatures considered in that investigation were 900°C and 1300°C. The mass loss was investigated by using crucibles made of platinum and rhodium, where the original mass of platinum and rhodium was about 150 g and 33 g respectively. The mass loss was measured at time intervals of 2 h for a total period of 30 h. The significance of the results from this investigation was that at 900°C the mass loss of the platinum and rhodium crucibles was found to be negligible. However, at 1300°C it was found that the mass loss of both platinum and rhodium increased linearly with time. This measurement showed that volatile oxides are formed and vaporise at high temperatures, and that these are formed above some threshold temperature (3, 11).

Jehn (5) also observed a linear dependence of the mass loss with elapsed time, though it was noted in that study that in other experiments extended over “extremely long time periods” a marked increase in the mass loss rate had been observed for platinum at 1300°C after about 60 h (12) and 400 h (13); no explanation was given.

One difficulty with previous measurements of the rate of mass loss of platinum and platinum-rhodium alloys is the very wide dispersion of results, which vary by several orders of magnitude. This may be due in part to the wide variation of geometries of the test conditions, and the wide variation of environments (such as ambient gas or vacuum). Furthermore, none are applicable to typical thermocouple formats, which comprise thin noble metal wires inserted in narrow ceramic (typically alumina) bores. In addition, no attempt has been made to link the rate of mass loss of the wires to thermocouple drift performance. The aim of the current investigation was to perform such mass loss measurements in a manner that is directly applicable to thermocouples, to better inform thermocouple drift models based on evaporation of platinum and rhodium oxides. Wires of Pt-30%Rh, Pt-6%Rh and platinum were investigated in ambient atmospheric conditions, within a typical alumina insulation tube used for thermocouples to make the test as realistic as possible in the context of thermocouple usage. A temperature of 1324°C was selected for the tests, because this is the melting temperature of the cobalt-carbon alloy, which is a well-established temperature reference point (a so-called ‘fixed point’) for calibrating thermocouples, and is used for in situ periodic calibration during long-term experimental thermoelectric drift tests.

In this article, the method is described, followed by a presentation of the results and an assessment of the associated uncertainties. The results are compared with thermocouple stability measurements where there are some intriguing qualitative similarities. Finally, some conclusions are drawn.

Two identical wires were used for each type of wire to investigate the change of mass, where one wire was used as the control (kept at room temperature) and the other was the test wire exposed to high temperatures. This allows the change in mass to be calculated by comparing the masses of both wires. Measuring the mass change in this way reduces the influence of calibration drift and scale linearity of the mass balance. The wire diameter for all wires was 0.5 mm. The length of the platinum wire was 60 cm, while the Pt-6%Rh and Pt-30%Rh wires were 70 cm long.

Each of the wires was thoroughly cleaned using acetone and distilled water in order to remove any surface contamination prior to making mass measurements and heating. The wire masses were then measured using a Sartorius Supermicro S4 balance (Sartorius AG, Germany) with a resolution of 0.1 μg and a standard deviation of less than 1 μg. The measurement uncertainty of the mass comparison performed was approximately ±25 μg, largely due to random fluctuations in the readings. The balance was placed in a laboratory at a constant temperature (20°C ± 1°C), and the apparatus was kept in an isothermal enclosure to prevent the air from the surroundings from interfering with the measurements.

Prior measurements were made to determine how the surroundings affect the mass measurements using two scrap platinum wires. One wire was kept as a control, where the scales were tared, and the other was the test wire (which in subsequent measurements would be the one exposed to high temperatures). The air conditioning in the laboratory caused short-term temperature fluctuations and air currents which manifested as long stabilisation times and noise in the mass measurements, so the thermal isolation of the balance was progressively improved until the stabilisation time and random fluctuations were both minimised.

To facilitate measurement on the balance pan, the wire was coiled around a plastic former to minimise contamination of the surface; the former was removed before weighing. Mass measurements of the test wires were made by comparison with the corresponding control wire, which was used to tare the balance. The mass of the control and test wires was measured three times to determine the average mass and to account for, and characterise, the random fluctuations in the test readings.

For each type of test wire used, three single alumina tubes of length 40 cm and diameter 3 mm, with bore diameter approximately 1 mm, were used. The wire was threaded into the bore to emulate the thermocouple format. The assembly was then placed in a thermal annealing furnace, which was maintained at 1324°C with stability of approximately ±1°C. Before beginning the experiment, the alumina tubes were maintained at 1500°C in air for 6 h to drive off volatile impurities, in order to minimise contamination of the test wires.

The platinum and platinum-rhodium test wires were placed into the alumina tube bores by straightening the test wires (which are coiled for the mass measurements), while the control wires were kept coiled and stored in clean plastic bags.

The bores containing the test wires were placed in the furnace for designated intervals before being removed at the end of each interval and coiled again to perform the mass measurements.

The uncertainty of the mass measurements was assessed by considering the individual contributions listed in Table I, which were summed in quadrature in a manner consistent with the International Organization for Standardization (ISO) Guide to the Expression of Uncertainty in Measurement (GUM) (14). The two statistical contributions, i.e. repeatability and reproducibility, were assessed by considering the standard deviation of repeated weighings in situ and of re-weighing the same item (i.e. removing from the balance and replacing) several times. The resolution, accuracy of the weights used to calibrate the balance, linearity and temperature of the balance reading were obtained from the manufacturer’s specifications. For general applicability of the results, the mass loss was expressed per unit surface area (mg cm^–2); this was determined by dividing the mass loss values by the surface area of the wires. This yields an uncertainty (k = 2) of ±0.0027 mg cm^–2 for the platinum wire and ±0.0023 mg cm^–2 for the Pt-6%Rh and Pt-30%Rh wires.

It was found that all the wires progressively lost mass during high temperature exposure. Figure 1 shows the mass loss per unit surface area as a function of time of exposure to high temperature. It can be seen that the mass loss is a linear function of elapsed time, at least up to about 150 h. Beyond that, however, a marked departure from linearity is seen, which suggests an acceleration of the mass loss per unit surface area. Intriguingly, previously reported very high precision drift tests of a set of different platinum-rhodium thermocouples (shown in Figure 2) show that they all exhibit a marked change in drift rate at about 150 h as well; beyond about 150 h the drift rate is constant for many hundreds of hours (15). The Pt-30%Rh appears to lose mass fastest, which is what might be expected since this wire contains more rhodium, and the vapour pressure of RhO₂ is higher than that of PtO₂ at this temperature (16).

Fig. 1.

Fig. 2.

The fact that the mass change up to about 150 h is linear for the reported measurements is in qualitative agreement with previous results (5, 10, 11). However, the rate of mass loss of platinum, using the data of Jehn (5) and Phillips (10), was about 0.0049 mg cm^–2 h^–1 and 0.0048 mg cm^–2 h^–1 respectively at 1350°C. In the current study, the rate of mass loss up to 150 h was about 0.0006 mg cm^–2 h^–1, which is about an order of magnitude lower. This may be because in the present study the wire is enclosed in quite tight-fitting bores to emulate their use in thermocouples, so the reduced rate of mass loss may be due to the establishment of saturated vapour pressure above the wire (inhibiting further evaporation), or reduced oxygen present as the surface oxidation proceeds, or deposition of the oxide vapour or all of these effects. This is in contrast to the previous studies cited, where the samples were bulk (for example, crucibles, meshes, coupons) and the surface was exposed to the surroundings rather than enclosed in tight-fitting bores.

The mass losses of platinum, Pt-6%Rh and Pt-30%Rh wires, commonly used for thermocouples, were considered as a function of elapsed time at 1324°C. The wires were placed in thin alumina tubes to emulate the thermocouple format. It was found that the mass loss of the three wires increases linearly with elapsed time, consistent with other investigations, up to an elapsed time of about 150 h, but after that, a marked acceleration of the mass loss is observed.

In the latter case, it is thought that, because the whole length of the thermocouple which develops an output is exposed to the full range of temperatures from room temperature to the measurement junction temperature, the behaviour in the first 150 h (17) arises from early development of microscopic ordering in the platinum-rhodium alloys (during exposure to temperatures below 600°C). This is followed by a slow coating of the platinum-rhodium leg by rhodium oxide (during exposure to temperatures between 600°C and 950°C) (18–22). This is difficult to apply directly to the mass measurements because the all of the wire was held at 1324°C so the oxidation effects at lower temperatures do not occur, but it is conceivable that comparable ordering and coating behaviour is taking place during the slow increase and decrease of the setpoint temperature of the furnace. It is also possible that the mass loss simply varies non-linearly over the whole duration: the motivation for expressing the data in terms of a discontinuity at about 150 h duration has been guided by the previously observed marked increase in the rate of mass for platinum at 1300°C after about 60 h (12) and 400 h (13); and also by the rapid change in thermocouple drift rate at about 150 h. To draw a definitive conclusion on this point, more detailed measurements would be needed, perhaps with the wire exposed to the entire temperature range from ambient to 1324°C.

The rate of mass loss was approximately an order of magnitude less than that observed in previous studies, and the enclosed nature of the wires in the alumina insulation tubes is thought to be a cause of this. Previous high precision studies have shown that a change in the drift rate after about 150 h at 1324°C is also observed in the thermoelectric drift of a wide range of platinum-rhodium thermocouples, and the current results were compared with those studies. The mass loss was greatest for Pt-30%Rh, followed by Pt-6%Rh, then platinum.

Components in many different applications at high temperature and in corrosive environments require materials with excellent high temperature mechanical properties and chemical resistance. Aerospace applications represent an extremely challenging field, for the development of new materials, and for the improvement of the existing ones (1). NBSAs with the Ni₃Al intermetallic compounds as the main strengthening phase have been widely used in high-temperature components such as aeroplane engines, industrial gas turbine blades and modern industry fields. After decades of development, the working temperature of NBSAs is about 1100°C (2) and can reach up to 1150°C (3, 4). Recently, a new platinum-modified nickel-base alloy with exceptional high temperature stability has been identified (5, 6). Coarsening studies conducted reveal unusually high volume fractions of morphologically stable γ’ precipitates up to 1200°C, which suggests that the alloy would have excellent performance as a bond coat or single crystal blade. A further increase in the operating temperature of the gas turbine will improve the combustion efficiency, reduce fuel consumption and CO₂ emissions, and leads to higher thrust values (7, 8). NBSAs operating temperature is approximately 85% of their melting point. An increasing interest has been shown in developing new alloys based on materials with higher melting points with similar structure to that of NBSAs and capable of being used at 1300°C (9, 10).

Potential candidate materials that may replace NBSAs or iron-based superalloys such as PM2000 mainly include intermetallic compounds, ceramics and ceramic matrix composites, refractory metals and high melting point platinum group metals (11). Some high melting point intermetallic compounds have high-temperature strength, lower diffusion rate and excellent corrosion resistance. However they lack plasticity or fracture toughness at room temperature (12, 13). Components based on ceramics and ceramic matrix composites have high-temperature strength, creep resistance, oxidation resistance and corrosion resistance which are attractive for their potential use in gas turbine engines. Unfortunately, low fracture toughness and brittle behaviour usually associated with ceramics are problems for high-temperature applications. Refractory metals and their alloys have high melting temperatures and for this reason researchers are considering the possibility of using them in the hot parts of gas turbines to replace NBSAs. These refractory metals and alloys lack sufficient oxidation resistance which limits their practical application (14). Superalloys based on platinum group metals (platinum, iridium, rhodium) show extremely strong chemical stability and two-phase structure (face-centred cubic (fcc)/L1₂), which makes this group of alloys a potential candidate to be developed as high-temperature materials for next-generation gas turbines (15–22). However, the main weaknesses of most iridium-based and rhodium-based refractory superalloys are brittleness, high cost and high density.

Unlike iridium and rhodium, platinum has become an essential high-temperature material in special applications. It has a higher melting point than nickel (platinum = 1769°C, nickel = 1453°C), better oxidation resistance, corrosion resistance, chemical stability and does not require coating protection when used at high temperatures (23, 24). Platinum-based alloys have excellent mechanical properties such as high creep strength and ductility, which make them have application potential in the fields of chemical engineering, space technology and glass industry (24, 25). At this time, the research on platinum-based superalloys includes solid solution strengthened, dispersion strengthened and precipitation strengthened alloys as well as platinum group metal compounds (26). Current usage is restricted to solid solution strengthened alloys and dispersion strengthened alloys, the latter being classified as part of the group of composite materials. Solid solution strengthened platinum-based alloys are a family of alloys which have been researched and developed for some time. The compositions and preparation process can be said to be more mature.

All transition group elements have considerable solid solubility in platinum. The elements near platinum in the periodic table form a continuous solid solution with platinum and have different degrees of solid solution strengthening effect on the platinum matrix. At high temperatures, ruthenium, iridium and rhodium have higher tensile strength than platinum and palladium. The high temperature durability and creep rupture strength of iridium are also much higher than that of Pt-Rh alloys. Ruthenium, iridium, rhodium and palladium have become the main solid solution strengthening elements. According to the relationship between stacking fault energy and creep rate, ruthenium and iridium have the largest solid solution strengthening effect on platinum, followed by rhodium, and palladium with the smallest effect (27).

Currently, the research and development on platinum-based solid solution alloys mainly include binary alloys such as Pt-Rh, Pt-Ir, Pt-Ru, Pt-Ni, Pt-W and ternary alloys such as Pt-Pd-Rh and Pt‐Rh-Ru (28). The properties of Pt-Rh alloys are the most stable: an increase in rhodium content leads to higher temperature durability, extended creep life and decreased creep rate (29). However, the improvement of mechanical properties decreases when rhodium content exceeds 30 wt%. In addition, machinability is significantly worse at these levels of rhodium content. The high temperature durability, creep life and creep rate of Pt-Ir alloy are better than the Pt-Rh alloy. On the other hand, Pt-Ir alloys tend to have higher weight losses in oxidising environments above 1100°C due to the selective oxidation of iridium after prolonged exposure to such atmospheres. In addition, for solid-solution strengthened Pt-Rh alloys, the coarsening of crystal grains at high temperatures will lead to reduced alloy strength and premature failure of components.

In order to improve the high-temperature mechanical properties of platinum-based alloys, oxide dispersion strengthened (ODS) alloys with ZrO₂ or Y₂O₃ as reinforcing phase have been researched and developed (30, 31). The fine oxide particles dispersed in the platinum matrix can stabilise the grain boundaries, prevent the movement of dislocations and improve the high temperature fracture strength. However, these ODS alloys have great brittleness, crack sensitivity and cannot withstand severe temperature changes. A particular type of ODS alloys has been developed, namely dispersion hardened platinum (DPH) alloys (32). These alloys are reinforced by dispersion of oxides formed inside the alloy via an internal oxidation process of pure oxygen-reactive elements (i.e., elements with high affinity to oxygen such as cerium, yttrium, scandium and zirconium). These elements are added to the melt and their oxidation is subsequently obtained by appropriate treatments. The stress-fracture strength of DPH alloys is further improved with respect to solution strengthened alloys. For instance, the stress-fracture strength of DPH platinum is higher than that of Pt-10Rh alloy (33), with good plasticity. However, the production process of all the dispersion strengthened alloy families is complicated and problems may arise during welding operations. The strength of the welded joint tends to drop due to excessive formation of oxide particles during welding (24).

Precipitation strengthening is the main strengthening mechanism of platinum-based superalloys currently under research and development. These alloys show higher temperature strength than solution strengthened and dispersion strengthened alloys at 1300°C (34–39). However, high density and cost are the major drawbacks to the use of platinum, but it is likely that the platinum-based alloys can be used for the highest application temperature components (10, 40). Due to the excellent properties of oxidation and corrosion resistance, the Pt-Al system superalloys could have potential as coatings on NBSAs or other substrates (10, 22, 41). Their density can be slightly reduced by adding suitable light alloying elements, while the high performance and recyclability of platinum-based alloys make up for the high price (42). This paper summarises the research status and progress of platinum-based superalloy materials. Firstly, we introduce the composition and structural optimisation design of platinum-based superalloys, the structural characteristics and evolution of Pt‐Al-based ternary, quaternary and multi-element superalloys, and their mechanical properties, oxidation and corrosion resistance behaviours. The strengthening mechanisms, the relationship between oxidation, corrosion and alloy composition have been analysed and the results will be presented and discussed. Finally, further research and application prospects of platinum-based superalloys are analysed and discussed.

A large number of new alloys have been researched and developed based on strong demand for higher working temperature and high temperature resistant structural materials in the aerospace field. The goal of the research is to seek a material with better high-temperature mechanical properties (tensile, fracture, creep and thermomechanical fatigue properties) and environmental stability (resistance to high-temperature oxidation and hot corrosion) than nickel-based superalloys (43). Inspired by the successful experience of precipitation strengthening obtained in the γ matrix (fcc structure) of nickel-based superalloys, much effort is now put into the research and development of platinum-based superalloys with similar structures to the γ/γ’ system found in nickel-based superalloys (7). Research on platinum-based superalloys began at the end of the 20th century. The initial work was mainly on structure and composition design, including the formation of a Pt₃X (γ’) precipitation strengthening phase and the selection of solid solution strengthening elements.

2.1 Second Phases and Possible Reinforcement

2.2 Alloy Composition Design

3.1 Structural Characteristics of Ternary Alloys

In order to obtain an effective strengthening effect on platinum-based ternary alloys it is necessary to determine the low-temperature structure of the Pt₃Al phase in the representative binary alloy. Wolff (22) used an electric arc furnace (EAF) to smelt the Pt-12Al (at%) alloy and the material was then subjected to 96 h solution annealing and subsequent ageing at 1350°C. Figure 2 shows some structural features of the Pt-12Al alloy. It can be seen that Pt₃Al precipitation phase exhibits strong directional cubic alignment and the dispersed sub-micron nature of the precipitation phase exhibits a typical bimodal size distribution. This special-orientation geometrical configuration is a highly coherent phase which has low mismatch strain. The Pt₃Al-γ’ phase additionally has a low-temperature variant and a lath or twin structure.

Fig. 2.

TEM bright-field image of cuboidal Pt₃Al precipitates in Pt-12A1 alloy after solution annealing and subsequent ageing at 1350°C (22). The magnification is 10,000 times

The maximum volume fraction of the Pt₃Al phase is limited to about 30% due to the maximum solubility of aluminium in platinum. The volume ratio of the precipitation phase in nickel-based superalloys can be as high as 70–80%. Additional information about the microstructural features of Pt₃Al can be inferred from Figure 3 where a transmission electron microscopy (TEM) bright-field image microstructure characteristic of the precipitate phase in an alloy with slightly higher aluminium content, namely Pt-14Al (at%) alloy (54). It can be seen from Figure 3 that the martensitic transformation has occurred in the platinum matrix leading to the formation of a clear stacked sheet or plate-like structure. Electron diffraction analysis shows that the stacked plates are twinned with each other, the [001] directions of adjacent tetragonal structure D0’c single packages are perpendicular to each other, and the twin planes are (112) planes. The [001] direction of D0’c is parallel to the <001> direction of the cubic matrix.

Fig. 3.

Bright-field TEM image of a Pt₃Al precipitate in a Pt-14Al alloy. Stacked plates or laths e.g. P1 and P2; M = platinum matrix. The arrows indicate individual platelets between stacked plates. Reprinted from (54), Copyright 2007, with permission from Elsevier

Adding transition metals such as nickel, titanium, chromium, ruthenium, iridium, rhenium and tantalum to Pt-Al alloys can improve the strengthening effect of γ’(Pt₃Al) precipitates, thermal stability, solid solution strengthening of the matrix and overall properties. Several typical ternary alloy microstructures are shown in Figure 4. The two-phase structure of Pt₃Al and platinum solid solution of all alloys has been confirmed by X-ray diffraction (XRD) experiments (55). The alloys with nominal compositions (at%) of Pt-14Al-3Re, Pt-14Al-4Ti, Pt-14Al-4Ta and Pt‐14Al‐4Cr were all characterised by a microstructure consisting of primary Pt₃Al surrounded by a fine two-phase eutectic-like mixture of a (platinum) matrix and fine particles of Pt₃Al. The proportion of primary Pt₃Al in Pt-14Al-4Ti, Pt-14Al-4Ta and Pt-14Al‐4Cr alloys is between 40% and 50% and in Pt‐14Al‐3Re alloy it is about 25% according to optical microstructure analysis. The fine martensite-like lamellar structure was observed in Pt-14Al-3Re (Figure 4(a)) and Pt-22Al-2Ru (Figure 4(e)) alloys. This means the Pt₃Al phase has transformed from the high-temperature cubic structure (L1₂) to the low-temperature tetragonal structure (D0’c). In the Pt-14Al-4Ti and Pt-14Al-4Ta alloys, it seems that the high-temperature Pt₃Al phase is formed and maintained. This demonstrates that the third metal elements titanium and tantalum can stabilise the L1₂ polymorph. It is important to observe that an alloying element must enter the Pt₃Al phase in order to prevent the low-temperature transformation of the Pt₃Al phase (55). Biggs (56) and Hill (57) have also shown the possibility for other alloy third components (nickel, titanium or chromium) in Pt-Al alloys to stabilise the cubic (L1₂) structure of Pt₃Al.

Fig. 4.

Scanning electron microscopy-backscattered electron micrographs (SEM-BSE) showing the microstructures of the Pt-Al-X alloys: (a) Pt-14Al-3Re; (b) Pt-14Al-4Ti; (c) Pt-14Al-4Ta; (d) Pt-14Al-4Cr; (e) Pt-22Al-2Ru. Reprinted from (55), Copyright 2001, with permission from Elsevier

Hill et al. (58) studied the microstructure and lattice mismatch of Pt-Al-X alloy systems with stabilising elements titanium, chromium, tantalum, ruthenium and iridium. Figure 5 shows TEM images of the typical two-phase microstructure. All the precipitated phases show a bimodal or even trimodal size distribution. Titanium, chromium and tantalum elements enter into the Pt₃Al phase to stabilise the cubic L1₂ structure, the precipitation phase has a cubic appearance with no clearly discernible internal structure (Figure 5(a)). On the contrary, when ruthenium and iridium enter the platinum matrix, the Pt₃Al precipitation phase transforms into a D0’c structure, and presents a band-like structure with alternating light and dark distributions (Figure 5(b)). Under higher magnification, it was confirmed that the lamellar structure in the precipitation phase of the D0’c tetragonal structure should belong to the twin structure (54).

Fig. 5.

Typical TEM images of the ~Pt₃Al precipitates in Pt-Al-X alloys (X = chromium, iridium, ruthenium, tantalum or titanium), with letters indicating the different size ranges (where P = primary; I, T = intermediate; S = secondary): (a) L1₂ precipitates stabilised by chromium, tantalum and titanium additions. The inset shows the selected area diffraction (SAD) pattern, confirming the L1₂ structure; (b) D0’c precipitates stabilised by iridium and ruthenium additions. Reused from (58), Copyright © 2001 by The Minerals, Metals & Materials Society. Used with permission

The lattice misfits between the matrix and precipitate phase in the Pt-Al-X alloy system at room temperature and 800°C were measured respectively by XRD (58). The (220), (211) and (112) diffraction peaks are used to obtain the lattice constants of the platinum solid solution matrix, L1₂-Pt₃Al and D0’c-Pt₃Al precipitate phases, respectively (a_matrix and a_ppt represent the lattice constants of the platinum solid solution matrix and Pt₃Al precipitation phase, respectively). Then, the Lattice misfits δ between the precipitation phase and the matrix are calculated by Equation (i), and the results are listed in Table III. It can be seen that the degrees of mismatch for all alloys are negative and the difference in mismatch degrees at different temperatures is very small. Besides alloys containing ruthenium, the lattice misfit increases at high temperature and the cubic L1₂ structure (alloys containing chromium, tantalum or titanium) has a lower degree of mismatch than the D0’c structure (alloys containing iridium, ruthenium).

(i)

Table III

Lattice Misfits Between Precipitates and Matrix for Selected Pt-Al-X Ternary Alloys (58)

		Room temperature			800°C
Alloy	Pt3Al type	amatrix, nm	appt, nm	δ	amatrix, nm	appt, nm	δ
Pt-10Al-4Cr	L12	3.9022	3.8741	–0.0072	3.9390	3.9103	–0.0073
Pt-10Al-4Ir	D0’c	3.8983	3.8507	–0.0123	3.9246	3.8747	–0.0128
Pt-10Al-4Ru	D0’c	3.9001	3.8530	–0.0121	3.9349	3.8967	–0.0098
Pt-10Al-4Ta	L12	3.8941	3.8682	–0.0067	3.9246	3.8961	–0.0073
Pt-10Al-4Ti	L12	3.8921	3.8642	–0.0072	3.9246	3.8961	–0.0073

3.2 Mechanical Properties of Ternary Alloys

3.3 Oxidation Behaviour of Ternary Alloy

Early studies on platinum-based superalloys mainly focused on the addition of alloying elements to improve oxidation resistance and ensure the γ’ phase has a stable L1₂ structure. Research shows that the volume fraction of the precipitation phase reaches only about 30% no matter how the heat treatment process is optimised. Hence it is difficult to obtain the desired strength of the alloy (60). Subsequent work mainly focused on the Pt-Al-Cr ternary alloy system. Adding nickel, ruthenium and other alloying elements to form quaternary and multi-element platinum-based superalloys is expected to further improve the microstructure and to enhance the mechanical properties and oxidation resistance. Compared with steel, nickel-based alloys and aluminium alloys, the experimental data and phase diagrams of platinum-based alloys are relatively lacking. In order to develop quaternary and multi-element platinum-based alloys, relevant research institutions in South Africa, Germany and the UK have collaborated to establish Pt-Al‐Ru (61) and Pt-Cr-Ru (62) ternary system and Pt‐Al‐Cr‐Ni (63) quaternary system alloy databases by experiment and first-principles thermodynamic calculations.

4.1 Pt-Al-Cr-Ni Quaternary Alloy

4.2 Pt-Al-Cr-Ru Quaternary Alloy

For materials used in high-temperature environments, the focus is usually on high-temperature strength (74). However, the creep, oxidation and corrosion resistance of alloys are also important (75). Increasing operating temperature will lead to continuous corrosion. Therefore, it is necessary to evaluate the corrosion resistance of high-temperature materials during the selection process (1, 76). For nickel-based superalloys, the high-temperature strength of the alloy is improved by increasing the content of aluminium and reducing the content of chromium, but the alloy is more sensitive to high-temperature corrosion, and it is necessary to develop and introduce protective coatings (77).

Since platinum-based alloys show excellent performance in various high-temperature applications such as glass manufacturing and corrosive substance processing, platinum-based alloys can be used to solve problems encountered in the aerospace industry (29, 78). Platinum-based superalloys are relatively new high-temperature materials, and there are very few literature reports on their corrosion properties. Fuel or intake air usually contains impurity elements such as sodium, sulfur and vanadium, which may form molten salt corrosion products such as Na₂SO₄, NaCl and V₂O₅, which in turn may lead to high temperature hot corrosion (HTHC) of materials (22, 79). Hot corrosion caused by molten salt or corrosive gas accelerates the oxidation degradation of high temperature materials, adversely affects the mechanical properties of the alloy and shortens the service life of high temperature components. There are two types of hot corrosion in nickel-based superalloys: Type I and Type II. Type I hot corrosion is also known as HTHC and usually occurs in the temperature range 850–950°C. Type II hot corrosion is also known as low temperature hot corrosion (LTHC) and generally occurs in the temperature range 650–800°C (79). Type I or HTHC is the main corrosive process in aircraft gas turbine engines.

Maledi et al. (80) studied the accelerated corrosion behaviour of five platinum-based superalloys in analytically pure anhydrous Na₂SO₄ at 950°C and compared them with NBSA with 1.25 μm thick Pt₂Al coating protection or no coating. The experimental results are listed in Table VI (76). The corrosion kinetics of nickel-based and platinum-based superalloys for the first 50 h are shown in Figure 18 and Figure 19, respectively. Due to the protective oxide layer formed on the surface of the platinum-based alloy, the weight gain associated with corrosion is very small. On the other hand, uncoated NBSA form oxides in the initial stage of corrosion, leading to increased mass. After further exposure to the corrosive environment, non-protective oxides form and cause catastrophic corrosion damage. Although the coated NBSA has better corrosion resistance than the uncoated alloy, it still degrades prematurely compared to the platinum-based superalloy. Experiments show that in the molten Na₂SO₄ environment, platinum-based alloys show superior corrosion resistance compared to both coated and uncoated NBSA. Figure 20 shows the protective layer with strong adhesion formed on the surface of the Pt-10Al-4Cr alloy. The surface protective layer morphology of the Pt-10Al‐4Ru and Pt-11Al-3Cr-2Ru alloys is similar (80). Although the morphology of the surface oxide layer of platinum-based superalloys containing chromium or ruthenium appears porous, their resistance to hot corrosion is higher than that of the platinum-based superalloys containing cobalt. NBSA was subjected to penetration of sulfur under the surface oxide layer, leading to formation of chromium and nickel sulfides. This alloy showed the worst resistance to sulfidation and hot corrosion. Resistance to sulfidation was also the subject of studies in the experimental work of Potgieter et al. (81). The results of the study are shown in Table VI. The tests were performed in a 0.2% SO₂-N₂ mixed atmosphere and research conclusions were similar to those obtained by Maledi et al. (80).

Fig. 18.

Fig. 19.

Fig. 20.

More than 20 years of research and development have yielded interesting results for platinum-based superalloys. The results obtained so far are only the tip of the iceberg of a very interesting and topical subject characterised by excellent prospects for future use. Optimised design of alloy composition, microstructure characteristics, mechanical properties and oxidation corrosion behaviour have been achieved. Among the systems studied so far, the Pt-Al-Cr-Ru alloy system has been selected and optimised for excellent performance. The ultimate goal is to develop platinum-based superalloys for application in industrial fields such as gas turbine engines. However, this is a competitive market that is difficult to penetrate with new materials. One possibility would be to exploit the possibility for platinum-based alloys to be used at temperatures 200°C higher than for NBSA (82). During the transition period, platinum-based superalloys could be used in other fields, such as castings, powder metallurgy products and coatings to accelerate their final use in gas turbine engines.

At present, the composition and structure design of platinum-based superalloys mainly follows the successful experience of NBSA development. The melting point of platinum is 316°C higher than that of nickel, but the difference in melting point between platinum-based superalloys and nickel-based superalloys is less than 150°C. The main reason for this discrepancy is the addition of low melting point alloying elements (such as aluminium) which reduces the melting point of the resulting alloy to about 1500°C. The advantage of platinum’s high melting point has thus not yet been fully exploited. The development cycle of a new generation of superalloys is very long and the material cost is relatively high for precious metals. The optimisation process for the design of new platinum-based superalloys could be accelerated with the help of material genome research concepts to further increase the alloy’s temperature tolerance and reduce research and development costs.

Platinum-based superalloys can still be considered a brand new alloy system when compared to, say, NBSA or stainless steels. The accumulation of fundamental data such as phase precipitation mechanisms and alloy properties is far from being complete. There has been little research on the influence of manufacturing processes (for example, precision casting, directional solidification and single crystal preparation) on the formability and mechanical properties of the alloys. There has similarly been insufficient verification and assessment of performance under actual use in typical environmental conditions. To accelerate bringing these materials to market, the level of research and development on platinum-based superalloys needs to be improved urgently.

The initial report (2) of fluxing of liquid Pd₄₀Ni₄₀P₂₀ with anhydrous liquid B₂O₃ showed that this treatment facilitates reliable production of fully glassy samples with larger cross-section. Although other agents such as soda-lime glass (3) have been used, B₂O₃ has remained the flux of choice for palladium-based and, later, platinum-based systems (4–15). While fluxing is an additional processing step, it is used because the effects can be dramatic: it can improve GFA, characterised as a reduction in R_c, by one or more orders of magnitude (6, 11, 16, 17). In a Pd-(Cu,Ni)-P BMG, fluxing increases the onset time at the nose of the time-temperature-transformation (TTT) curve for crystallisation in the liquid from 130 s to 200 s (17) and has been critical in achieving BMG castings over 50 mm in minimum linear dimension (18, 19).

Fluxing works by reducing the influence of heterogeneous nucleants that would facilitate crystallisation on cooling the melt. These nucleants are generally considered to be dispersed particles, mostly oxides, in the melt, but fluxing may also prevent contact with nucleants at the melt surface. Fluxing works by preferential wetting of the particles, possibly followed by dissolution of the particles in the flux. The removal of oxide particles from the melt would reduce its oxygen content. Direct verification of this is rare but, in one study, an initial oxygen content of ~12 ppm (by weight) was reduced to ~5 ppm by fluxing (8). Following the initial work (2), it is usual to cycle the melt between solid and liquid states several times; crystallisation in early cycles may assist in driving inclusions to the surface of the alloy, facilitating their take-up by the flux.

Ideally, fluxing would have no effect on the composition of the liquid alloy itself. In particular, when using B₂O₃, none of the elements in the alloy should have a greater ‘oxygen affinity’ (11), i.e. more stable oxide (as quantified by the relevant chemical potentials), than the boron in the flux. Alloys of the systems Pd-(Cu,Ni)-P and Pt-(Cu,Ni)-P meet this condition. But B₂O₃ has also been used to flux Pd-Cu-Si (8): in this case, reaction with silicon in the alloy can reduce B₂O₃, leading to dissolution of boron into the alloy. The addition of boron is detrimental to the GFA and the overall effects of fluxing are therefore complex, varying markedly with the time-temperature profile used in the treatment (8, 11). Nevertheless, by optimising the fluxing of Pd-Cu-Si, its d_c can be substantially improved to 11 mm (20) and 15 mm (8). Studies of Pd-Ni-Si-P alloys suggest that composition changes inducing by fluxing with B₂O₃ can in some cases be beneficial for GFA (7, 17, 21) and lead to improvements in plasticity (22).

The remarkable improvements in GFA for ternary BMGs suggest that the fluxing should be considered in greater detail. The high m-fragility of, and the boron content within, Demetriou et al.’s 950Pt based BMG suggests a sensitivity to fluxing and boron pick-up (23), so enhanced fluxing procedures should be sought. Perhaps the effects of fluxing and overheating (to dissolve preexisting structures in the melt) should be revisited for alloys where it has been previously reported that fluxing does not affect GFA. It would be good to explore the use of alternative fluxes. B₂O₃ has been very widely used, not least because it remains liquid below the T_g of the alloys being processed. It may be possible to find multicomponent fluxes that remain liquid to similar temperatures, yet have greater resistance to chemical reduction by reaction with the alloy being processed.

In the literature, the action of fluxing has been described as ‘purification’ of the liquid alloy. While fluxing may clean an alloy by removing dispersed particles, there is no evidence that it can reduce levels of dissolved impurities. In contrast, as noted above, the alloy can be contaminated by reaction with the flux. In the production of BMGs, impurities can be detrimental to GFA. The most prominent example is the presence of oxygen in zirconium-based alloys, as briefly reviewed elsewhere (11). It is important to start with raw materials of ultra-low oxygen content and care must always be taken to minimise oxygen pick-up during processing. Various approaches have been adopted to reduce the oxygen content; these include oxygen scavenging with low addition levels of elements such as yttrium and scandium (24). In addition to the scavenging effect, these elements with large atomic diameter are likely to have microalloying effects that improve the GFA (24).

For palladium- and platinum-based glass-forming alloys, in contrast, there is little concern about oxygen as a dissolved impurity. Nevertheless, it remains desirable (in some cases essential) to use raw materials of high purity and to minimise pick-up of impurities during processing.

In conventional investment casting of platinum alloys, their high T_L (around 2000 K) limits the casting size (25) and introduces many issues such as reactions with the crucible, tarnishing and oxidation, plus significant casting shrinkage and porosity (as high as 5%) (26). While palladium jewellery alloys are cast at slightly lower temperatures (27), the further reductions in casting temperatures offered by BMG-forming compositions are still desirable. The low-lying eutectics of platinum-based and palladium-based BMGs offer an exciting opportunity for easier processing of precious-metal jewellery alloys.

The main challenges in casting BMGs arise from the high cooling rates required to avoid crystallisation and the high viscosity of the glass-forming liquid compared with the melts of usual crystalline alloys. In conventional casting techniques, good form-filling requires long casting times (i.e. slow filling), which are longer for more viscous liquids. For glass formation, techniques with low cooling rates (for example, investment casting), are not suitable or need adapting. In any case, the casting size is limited by d_c. The undesirability of cold-working BMGs (due to possible fracture or the formation of unsightly surface steps) requires items to be cast near to their final shape.

Nevertheless, for direct conventional casting, BMGs present many advantages when compared with conventional crystalline alloys and this remains a practical processing route for jewellery. The absence of a first-order phase transition on cooling to form a glass means that there is no substantial shrinkage during casting. Even if any crystallisation were to occur, these glass-forming liquids exhibit small crystallisation shrinkages (28).

Casting porosity is a particular problem for many platinum jewellery alloys (26), motivating the introduction of hot isostatic pressing of castings. In palladium-based jewellery, 950Pd alloys suitable for high-quality investment casting are the focus of active research and development (29). For BMGs, low volume shrinkage (< 0.5%) on casting means near-net-shape casting can be achieved with few casting defects such as porosity and with high surface definition. Their lower casting temperatures also lead to lower thermal stresses during cooling to ambient temperature (30). Fewer polishing steps are required after casting due to the absence of crystallinity; ultimately, there can be an atomically smooth surface finish (31).

Unlike zirconium-based BMGs, which are already used in luxury goods markets (32), BMGs based on precious metals (gold, palladium and platinum) for jewellery have the considerable advantage of being processable in air (4, 30, 33–36).

The casting of gold-based BMGs for jewellery has been evaluated in detail (37–39). Similar information for platinum- and palladium-based BMGs is not as readily available, but their comparably high fragilities and larger d_c values (Table I) mean that their castability is similar, if not better. The only significant difference is the requirement to flux with B₂O₃ before casting. BMGs based on precious metals perform well compared with so-called ‘benchmark’ BMGs which have exceptional processability (32, 34).

Many BMGs can be made using methods such as tilt casting, suction casting and centrifugal casting. While methods such as suction-casting achieve higher cooling rates and higher-quality castings with low porosity (44), die-casting and tilt-casting techniques are scalable to high-volume production (45). High flow rates can induce crystallisation throughout the sample due to shear thinning of the liquid and should be avoided (46). As a result, methods such as centrifugal casting and continuous casting are difficult to apply to BMGs (38, 47, 48). Some heterogeneous nucleation of crystals is inevitable in contact with the copper mould (49).

For BMGs as a jewellery material, tilt-casting appears the most suitable conventional method. Although maximum cooling rates and form-filling are lower than for suction-casting, the method is suitable for industry and has been successfully used to cast samples in massive copper moulds in many of the studies presented here.

For the most intricate designs, a goldsmith uses investment casting. In this method, high form-filling is achieved by slow cooling. While slow cooling gives enough time for the melt to fill the entire mould (37), it is incompatible with the requirements for glass formation. To circumvent this problem, Eisenbart et al. developed a lost-metal mould casting technique, which is an analogue to investment casting for BMGs (39). In their technique, a wax pattern is produced and covered in a thin layer of conducting material. Electroforming is then used to deposit copper onto the wax pattern. The mass of copper deposited is typically five to 10 times the intended mass of the casting to ensure cooling rates high enough for glass formation (38). The wax pattern is then removed either by melting or chemically and the component is centrifugally cast. Finally, the copper mould is etched away to leave the final casting (39).

This technique could have widespread application for jewellery. It allows the formation of highly intricate shapes (Figure 1) not achievable with a dividable copper mould, and high form-filling (near 100%), while satisfying the requirements for glass formation. Furthermore, the electroforming and etching procedures mean that the copper is recyclable, and low-cost wax patterns can be easily shaped by an artisanal goldsmith or manufactured at industrial scale (38).

Fig. 1.

Thermoplastic forming (TPF), a method usually reserved for thermoplastic polymers and conventional oxide glasses, can be applied to BMGs. Aside from improved wear resistance, the ability to be so formed is perhaps the most desirable property of BMGs for jewellery. Current jewellery manufacture is laborious, requiring a highly skilled goldsmith. TPF of hallmark-compliant BMGs offers the potential for fast, economical, large-scale production of jewellery items alongside artisanal design innovations.

To date, BMGs have been successfully shaped via a wide range of TPF techniques such as blow-moulding (50–53), hot-stamping (54–57), extrusion (58–61), hot-rolling (62) and injection-moulding (63). Novel processing possibilities, potentially of use to jewellers for setting gems, include welding and joining methods designed to minimise crystallinity (64–66), as well as the incorporation of second phases when processing in the supercooled liquid region (SCLR) (67).

TPF has many advantages over the direct casting of BMGs and the casting of crystalline metals. It leads to even fewer casting defects than the direct casting of BMGs, since this isothermal process allows the relaxation of internal stresses and elimination of porosity. Porosity can be reduced to levels well below 0.2% (30). TPF can give a wide range of complex, thin-walled and even hollow shapes with high dimensional accuracy (51–53), as well as surface patterning with nanometre accuracy (55–57). Recent advances in stretch blow-moulding allow more complex shapes and plastic strains up to several thousand percent (52) compared with a few hundred percent for conventional blow-moulding (50). From a practical perspective, the substantially lower working temperatures and the ability to process platinum- and palladium-based BMGs in air (35) mean that equipment lifetime is extended, costs are lower and manufacture is safer.

Among the many advantages of TPF, the most important is the ability to produce fully glassy samples with dimensions exceeding d_c. This is possible since forming of the final shape and cooling to form the glass are decoupled (34). First, a glassy feedstock material (rods (62), pellets and granules (31, 38), plates (45, 67) or discs (50)) can be cast fully glassy. The feedstock can then be used for TPF to produce a larger component. The use of a pellet feedstock is common practice in polymer processing. It allows fast, easy and economical production since large quantities of simple-shape feedstock can be produced by straightforward casting, while complex shapes are introduced only at the final TPF stage. For Pd-Cu-Si based BMGs, as-cast granules have a SiO₂ surface layer. This is thicker than would otherwise be expected (68) due to the lower surface tension of silicon which promotes segregation to the liquid surface (69). As for gold-based BMGs, this SiO₂ should be removed before TPF to ensure adhesion between the feedstock particles (38, 66).

Since the BMG must be heated above T_g, TPF is possible only within a limited time-temperature processing window before crystallisation occurs (Figure 2). Crystallisation leads to embrittlement, a loss of thermoplastic formability and of many of the desirable properties associated with the glassy structure (36, 55).

Fig. 2.

While it would seem logical to perform TPF at temperatures just above T_g to maximise the available processing time, sharp-definition forming requires a low viscosity (below 10⁶ Pa s (45, 70)). As can be seen from a simple treatment of deformation in the SCLR using the Hagen-Poiseuille equation, total deformation increases as the temperature increases (35, 71). As the temperature increases by 20°C, the viscosity, η, decreases by one order of magnitude, but the time to crystallisation onset, t_c, decreases by a factor of three (36). Consequently, the maximum possible deformation increases, as has been confirmed experimentally (71, 72). The best TPF is achieved by heating the feedstock quickly to the highest temperature possible, shaping the low-viscosity supercooled liquid into the desired final product, then quenching back into the glassy state before the onset of crystallisation (35, 73). Given that a more intricate design takes longer to process, there is a desire to increase thermoplastic formability. Strategies being researched include optimisation of fluxing which can widen the SCLR (8, 11, 19), and multi-step thermoplastic processing (74).

By comparison with other BMGs, gold-, platinum- and palladium-based BMGs exhibit exceptional thermoplastic formability (Table I) due to their low T_g, high liquid m-fragility, large ΔT_x and large S (34, 36, 75). Values of F^scan, a parameter that correlates better with deformability during TPF but is more difficult to measure (34, 75), also show that among BMGs, those based on gold, platinum or palladium are the best BMGs for TPF (Figure 3) (75).

Fig. 3.

TTT curves for crystallisation of BMGs are helpful in comparing their stability during TPF. platinum-based and palladium-based BMGs show excellent stability against crystallisation at high temperatures within the SCLR (Figure 4). As a result, these alloys, alongside zirconium-based BMGs, have been widely used for many TPF experiments reported in the literature (Figure 5), from hot-embossing on the micro- and even nanometre scale (56), to blow-moulding and injection-moulding (52, 77). As examples, BMG watch casings can be made by TPF and machining; those using zirconium-based BMGs are now commercially available, while those using platinum-based BMGs have been made in research laboratories (40).

Fig. 4.

Fig. 5.

The intrinsic metastability of glass means it is vital to consider the effect of ageing and ultimately crystallisation, which can have a dramatic effect on properties (36, 73, 88–91). BMGs based on precious metals have low T_g’s compared to other MGs, where T_g can exceed 1000 K (92). At room temperature, BMGs based on precious metals are at a comparatively high fraction of T_g so the structure evolves on a shorter timescale. While this is of most concern for gold-based BMGs (T_g = 130°C (33)), it is also important to consider for platinum- and palladium-based BMGs.

In a broad range of temperature mainly below T_g, ageing of glass occurs via α and β relaxation. β relaxation can be associated with changes in chemical short-range ordering, leading to substantial changes in enthalpy. It is ultimately responsible for sub-T_g embrittlement of BMGs (93–95). α relaxation occurs predominantly near T_g or on much longer timescales (96–99). It is responsible for densification of the glass as it evolves towards the state of a deeply supercooled liquid. This densification, often described as a reduction in ‘free volume’ (100–104), occurs during long sub-T_g annealing, and is associated with hardening (94, 95).

Crystallisation can occur only with a level of atomic mobility equalling or exceeding that necessary for α relaxation. Mostly seen on annealing, crystallisation can also be induced by irradiation, deformation or dealloying. It is classified into polymorphic, eutectic and primary crystallisation (105–107). In the polymorphic and eutectic cases, the crystallised regions have the same composition as the original glass; as crystallisation proceeds, the glass composition stays unaltered. In that case, on annealing at a fixed temperature, the crystal growth rate is constant. In contrast, primary crystallisation involves solute partitioning between crystal and glass; the associated diffusion control tends to limit crystallite size and fine dispersions may be obtained (107). Crystallisation often occurs in many stages, via metastable phase combinations, towards the equilibrium phase mixture.

Partial crystallisation can lead to a dramatic increase in hardness (Figure 6). Despite suggestions that this is due to a phase-mixture effect (108), it can be due to solute enrichment of the glassy matrix, implied by rejection of solute from the growing crystalline phase (90, 109). The increase in hardness is predominantly attributed to the evolution of hard, brittle equilibrium phases, but we also note that crystallisation in Pt_49.95Si_6.4B₂₄Cu_16.65Ge₃ led to a rise in T_g and hardness of the remaining glassy matrix due to copper enrichment (from 570 HV to 750 HV) (88).

Fig. 6.

The equilibrium crystalline phases that form can have complex crystal structures. Although these complex structures may be regarded as beneficial, in that their slow growth should reduce R_C (110–112), they are hard and brittle. Partial crystallisation therefore leads to substantial embrittlement and a fall in fracture toughness (36, 55, 90, 91, 94, 95, 104, 108, 109). Cardinal et al. report that fracture toughness already halves as a result of 20% crystallinity (36, 74). Given the initially high fracture toughness of many platinum- and palladium-rich BMGs (23, 79, 82), substantial increases in hardness could be achieved, without the BMGs becoming too brittle. The potential advantages of partially crystallised systems, which can be regarded as ‘metallic-glass-matrix composites’ (MGMCs), should therefore not be ignored.

Surface crystallisation may lead to enhanced wear resistance. Kazemi et al. report that the surface hardness of their Pt-Cu-Si-B-Ge BMGs can increase to 800 HV while maintaining toughness in the rest of the sample (25). Surface treatments of this kind may be desirable, but the effect of surface crystallisation on other properties must also be considered.

From a scientific perspective, the crystallisation of BMGs is complex due to the simultaneous or stepwise crystallisation of many ordered phases. Their GFA, thermal stability and crystallisation rate are highly dependent on the fluxing procedure (17, 113), due to the effect of lower cooling rates and cleaning of the melt and on sample size due to the statistical probability of a heterogeneous nucleation site (114). On a TTT diagram (Figure 4), crystallisation of a glass during annealing is described by a C-curve (4, 17, 21, 113). The nose of the curve is often sharp since the crystallisation rate may transition over a narrow temperature range from being nucleation-limited to growth-limited (21, 113). Understanding the factors that affect crystallisation and crystallisation steps is particularly important for understanding an alloy’s suitability for TPF where crystallisation can readily occur. During use as jewellery, crystallisation is unlikely as the temperature is well below T_g.

Wu et al. compared the corrosion resistance of glassy and crystalline Pd₄₀Ni₄₀P₂₀ (120). The crystalline alloy (formed via vacuum annealing) performed two-to-three times better than the glassy alloy in a range of solutions, due to the presence of several inert phosphides and noble palladium-rich crystalline phases (120). These are the dominant equilibrium phases that form during the crystallisation in many Pt-P and Pd-P-based BMGs (113, 121) so similar corrosion resistance is expected. The formation of an inert surface layer appears critical to corrosion resistance (122). In the corrosion of Pd-Au-Ag-Si BMGs, passivation occurs due to the formation of a palladium-enriched surface layer with underlying silicon-rich glassy phases (123). The resulting anodic passive current density is lower than for SUS 316L stainless steel (123).

Copper-containing BMGs contradict these observations that crystallisation enhances corrosion resistance. Studies of Pd-Cu-Ni-P glasses and their crystalline counterparts in a variety of corrosive solutions show that the heterogeneity of the crystalline material results in an inhomogeneous passive film (124, 125). Subsequent pitting leads to accelerated corrosion (124, 125). During crystallisation, several equilibrium phases, Cu₃Pd, Ni₂Pd₂P, Cu₅Pd₃P₂ and an unknown quaternary phase, form via numerous metastable phases (126). The presence of less noble copper-rich phases amongst noble-metal inert phases leads to galvanic cells and the accelerated corrosion of the copper-rich phases (124). In corrosive chlorinated acidic solutions, copper and nickel dissolution is observed first, followed by the dissolution of palladium and phosphorus. This leaves behind a porous Pd-P network that is gradually dissolved after the outer layers of nickel and copper are removed (125). Dealloying may occur via dissolution of phosphorus followed by the formation of a nanocrystalline Pd-Ni solid solution as ligaments in an initial porous structure (127). Similar pitting was also observed in Pd-Cu-Si alloys (115, 127), suggesting that crystallisation is favourable for corrosion resistance only if the phases that form are all equally inert.

Dealloying in glasses occurs due to preferential leaching out of less noble elements. Their dissolution into the electrolyte allows the remaining noble-metal atoms to rearrange themselves into fine ligaments, resulting in a nanoporous surface (68, 128). Crystallisation due to the deviation of the composition from the glass-forming composition, contributes to the formation of ligaments by nucleation and growth of the crystals in the glassy matrix. Compared with dealloying in crystalline alloys, a finer structure is observed in glassy alloys (128). While the resulting nanoporous surface is desirable for catalysis and related applications (128–130), it has a degraded appearance unacceptable for jewellery.

Although few studies have compared tarnish resistance, Rizzi et al. reported that a Pd-Cu‐Si BMG incubated at 37°C in artificial saliva did not measurably tarnish while a gold-based BMG tarnished significantly (68). Tarnishing of palladium-based BMGs may not be as problematic as for gold-based BMGs (68) but, given the similarities in composition, it is not clear why this is so. Rizzi et al. note the existence of a SiO₂ layer on the surface of both alloys (68). In gold-based BMGs, this SiO₂ layer that forms on casting is an effective diffusion barrier, and prevents tarnishing, but it is quickly worn away and therefore it is not protective (68, 116, 117, 131).

Comparing gold-based BMGs and Pd-Cu-Si-based BMGs, the latter have a T_g that is more than 100 K higher. Atomic diffusion rates, closely linked to β relaxation (118, 119, 132–134), when compared at room or body temperature, are therefore likely to be significantly lower in Pd-Cu-Si-based BMGs. This might explain their slower tarnishing. Further studies comparing the behaviour of these alloys are likely to be significant in furthering the understanding of the tarnishing mechanisms of precious-metal-based BMGs.

For Pd/Pt-Ni-P based alloys, there appears to be no tarnishing problem. However, the presence of copper in these alloys impairs corrosion resistance, while replacing it with nickel leads to a lowering of GFA and raises issues of toxicity (36, 123, 135).

Further studies of corrosion and tarnishing of palladium- and platinum-based BMGs at body temperature and in simulated body fluids, such as on gold-based BMGs (38, 68, 116, 117), are required to fully characterise their suitability for jewellery, as well as for other applications such as in dentistry or as biomaterials (68, 123, 136).

Aside from the central issue of the conflicting requirements for high fineness and high GFA, BMGs for jewellery offer a wide range of desirable properties. In the as-cast state, they have hardness values that are, at the very least, challenging to achieve in conventional crystalline metals. With values exceeding 300 HV, components made from BMGs are suitable for watchmaking, potentially even as functional components, due to their unique combination of high resilience and hardness. BMGs in jewellery have excellent scratch-resistance, while more intricate, thinner, and even hollow designs are possible because of their high strength.

BMGs offer many processing advantages, both for casting and TPF. The lower casting temperatures of BMGs, alongside the absence of volume shrinkage due to crystallisation, result in relatively few casting-related defects (such as shrinkage and residual stresses), and in possibilities for near-net-shape casting and excellent surface finishes in the as-cast state. While the requirement for fast cooling limits the size, characterised as the critical casting diameter, d_c, of fully glassy samples, the attainable dimensions do exceed those required for most jewellery. The difficulties of form-filling, due to viscous melts and to high required cooling rates, have been resolved through new casting techniques. After casting, TPF of samples offers a wealth of new shaping opportunities. TPF may, therefore, enable new jewellery designs as well as economical, highly reproducible mass production.

The corrosion and tarnish resistance of platinum- and palladium-based BMGs are relatively unstudied. From the literature available, tarnishing does not appear to be a significant issue, while corrosion resistance is improved by the presence of noble phases on the surface and is superior to some stainless steels in aggressive environments. While surface coatings may appear to be a solution, inevitable wear leads to accelerated corrosion of any exposed underlying metal.

For platinum- and palladium-based BMGs, the main obstacle to their use in jewellery is the need to achieve fully glassy states with over 95 wt% platinum or palladium. Achieving such high fineness while alloying sufficiently to achieve adequate GFA appears, at best, challenging. The development of platinum- and palladium-based BMGs with precious-metal content exceeding 90 wt% and even 95 wt% (23, 79) does, however, give hope. Compositions that would give 950Pt and 950Pd BMGs are expected to show high liquid fragilities and therefore low GFA is expected.

In the search for higher precious metal contents, microalloying appears a key strategy, although its effects are far from fully understood. A deeper understanding of microalloying effects may assist in optimising GFA, but the restrictions imposed by hallmark-compliance mean that extensive microalloying can only be with lighter elements (to maximise the weight percent of platinum or palladium) and so other avenues should also be considered. MGMCs may offer opportunities to raise the precious metal content without sacrificing GFA or leading to a heterogeneous appearance (137).

Fluxing of the melt, to remove oxide particles that may act as heterogeneous nucleation sites, has a dramatic effect on GFA: optimised processing can increase critical casting thicknesses from the millimetre to the centimetre scale (8, 11, 18, 19). But treatment of some palladium- and platinum-based BMG-forming liquids with the usual B₂O₃ flux can lead to undesirable contamination with boron (23, 40, 79, 138). There is a clear need for the development of more stable fluxes and opportunity for optimisation of fluxing treatments.

Compared to 950Pt BMGs, 950Pd BMGs are even more challenging to achieve. From the compositions reported (Table III), further exploration of Pd-Cu-Si alloys should be considered. While Demetriou et al.’s palladium-based BMG exhibits exceptional mechanical properties, compositions such as Pd₇₉Cu₆Si₁₀P₅ (138) show higher GFA and higher weight fractions of palladium, due to a smaller degree of required alloying with metalloids. They, therefore, appear a promising area for further study in pursuit of hallmark-compliant BMGs. The Pd-Si binary system (specifically Pd₈₁Si₁₉) is one of the best binary glass-forming systems known (142, 143), due to a large energy barrier to crystallisation (142). As for Pd-P-based BMGs, these compositions have liquids of high m-fragility. With palladium content raised to 95 wt%, the m-fragility is likely to be even higher, resulting in reduced GFA. But, there are still many possibilities for a potential 950Pd hallmark-compliant BMG.

Lastly, for platinum- and palladium-based BMGs to be applicable in jewellery, they must maintain their lustre. Limited work has been carried out to assess how they may tarnish or corrode over time. Further studies of their performance in simulated body fluids at body temperature are necessary to evaluate their suitability. Such studies may also assist in the development of tarnish-resistant 18 karat gold-based BMGs.

Metal-metalloid alloy systems can be excellent glass-formers. This is attributed to the small metalloid atoms occupying interstitial sites between the dense-random-packed metal atoms. The strong bonding between metal and metalloid atoms further helps to give a stable dense packing (75). Pt-P and Pd-P binary eutectics are popular starting points for BMGs with further alloying additions, such as nickel and copper, leading to a substantial increase in GFA (76–78). From a topological standpoint, these additions help to form efficiently packed clusters. As proposed by Miracle et al., the form of these clusters depends on the relative sizes of the constituent atoms (79–82).

Following the discovery of palladium-based bulk glass formers by Chen and Turnbull (15), the Pd-Ni-P ternary eutectic has received particular attention. Partial substitution of nickel with copper leads to a dramatic rise in GFA. The heats of mixing for Pd-Cu, Ni-Cu and Cu-P pairs are more strongly negative than for Pd-Ni (76) and lead to a change in the dense packing of atoms in the viscous liquid (77). Further refinements in composition resulted in an alloy with d_c on centimetre scale and wide ΔT_x (83). The maximum GFA was later attributed to having the same chemical SRO of nickel and copper around phosphorus (78). These BMGs are particularly stable against crystallisation (84–88). When crystallisation eventually occurs in the SCLR, several ordered and complex crystalline phases are formed in a single cooperative transformation. The excellent GFA is attributed to the difficulty of crystallising the ordered Pd₃Cu phase. Compared to other BMGs, the nose of the crystallisation curve on the time-temperature-transformation (TTT) diagram (the temperature at which crystallisation is fastest) for Pd₄₀Ni₁₀Cu₃₀P₂₀ lies at a time 10 times longer than for the benchmark glass-former Vitreloy^TM 1 (85, 89). The time before crystallisation onset is, however, still short: for a Pd₄₃Ni₁₀Cu₂₇P₂₀ sample held just above T_g this is 10⁴ s (84, 89). Below T_g, times to the onset of crystallisation are orders of magnitude longer due to the dramatically lower atomic mobility.

While palladium-based BMGs have been studied since the 1970s, platinum-based BMGs are a comparatively recent development. In 2004, Schroers and Johnson reported two novel platinum-based BMG compositions, Pt_57.5Cu_14.7Ni_5.3P_22.5 and Pt₆₀Cu₁₆Co₂P₂₂ (28). These show high GFA when their melts are fluxed with B₂O₃ (d_c > 10 mm), good thermal stability (ΔT_x > 60 K), exceptional thermoplastic formability (S > 0.20), as well as being processable in air (28) (which is difficult for zirconium-based BMGs, for example).

For jewellery, Pt₆₀Cu₁₆Co₂P₂₂ is particularly attractive due to its high platinum content (satisfying the 850Pt hallmark) and the absence of skin-sensitising nickel (28, 41, 42, 45). On the other hand, Pt_57.5Cu_14.7Ni_5.3P_22.5 attracted much attention from researchers. Its plastic strain of 20% and fracture toughness of 80 MPa m^1/2 far exceed any previously reported values for BMGs (28, 72). Further studies have reported fracture toughness for both palladium- and platinum-rich BMGs as high as 200 MPa m^1/2 (73, 74) — a value comparable to many low-carbon steels widely used as structural materials.

This unique combination of high hardness and high plasticity has been linked to their high Poisson ratio (>0.4) in the glassy state, which itself is related to high m-fragility in the liquid state (72–74). High Poisson ratio corresponds to a low ratio of shear modulus G to bulk modulus B. The low value of G/B indicates that resistance to shear (proportional to G) is low compared to resistance to cavitation (proportional to B). While a typical BMG would fail by the operation of a single dominant shear band, leading to macroscopic plasticity below 1%, these BMGs flow by the operation of multiple shear bands leading to high macroscopic plasticity (72).

While high m-fragility of the glass-forming liquid is associated with desirable fracture toughness, it is also associated with reduced GFA (62, 87). The exceptional GFA of both Pt-P and Pd-P based BMGs is therefore surprising (72–74, 76, 77, 90, 91). As noted above, high GFA is typically associated with strong liquids (62). Fragile liquids do not aid glass formation through high viscosity (low atomic mobility), but high GFA may result also from a low thermodynamic driving force for crystallisation, or a high crystal-liquid interfacial energy (inhibiting crystal nucleation) (87, 92). Studies report that a low driving force for crystallisation stabilises Pd₄₃Cu₂₇Ni₁₀P₂₀ while high interfacial energy stabilises Pt_57.5Cu_14.7Ni_5.3P_22.5 (87). The nature of stabilisation helps to explain why fluxing has such a pronounced effect on the GFA of these alloys, since the presence of any heterogeneous nucleation sites, notably oxide inclusions, substantially lowers their resistance to crystallisation.

Measurement of crystallisation kinetics over the full temperature range from T_L down to T_g is helpful in understanding the mechanisms. A classical TTT diagram shows the times necessary for the progress of crystallisation in the supercooled liquid upon isothermal holding at each temperature. The times, for example for crystallisation onset, follow a C-curve in which the minimum time (at the nose of the curve) lies between T_L and T_g. At higher temperature than the nose, the kinetics is controlled by crystal nucleation, and at lower temperature by crystal growth. In most cases, for example, for a zirconium-based BMG-forming alloy, the C-curve is asymmetric with the temperature of the nose much closer to T_L than to T_g (62). In contrast, the C-curves for palladium- and platinum-based alloys are more symmetric (84, 87, 91). For Pd₄₃Ni₁₀Cu₂₇P₂₀ (89), for example, the nose lies roughly halfway between T_L and T_g. As the m-fragility of palladium-based BMG-forming liquids is relatively high (m>50 (93)), the kinetics nearer to T_L should be relatively accelerated, which, in the absence of other factors would impart greater asymmetry to the C-curve. The special feature of Pd₄₃Ni₁₀Cu₂₇P₂₀, and palladium- and platinum-based BMG-forming compositions in general, is thus identified as difficulty in crystal nucleation (89). Indeed, the nucleation is mostly possible only because of the influence of heterogeneities (89). This explains why fluxing (which can remove heterogeneous nucleation sites, notably oxide inclusions, Section 1.1, Part II (17)) can dramatically improve the GFA of these alloys.

Recent work on the structures of Pd-P and Pt-P based BMGs provides insight into the critical features required for high GFA (92, 94–97). Detailed studies show that the crystal-liquid interfacial energy in Pt-Cu-(Ni/Co)-P BMGs is three times that in kinetically stabilised zirconium-based BMGs (94). The dramatic improvement in GFA with overheating above T_L, to dissolve all preexisting structures in the liquid, provides further evidence that these glasses are stabilised by the barrier to crystal nucleation (94, 96).

In the search for BMGs with high weight fractions of platinum or palladium, understanding the liquid kinetics is imperative. Studies suggest that Pt-P based liquids with higher platinum content are more fragile (95, 97). This is relevant, given the desire in jewellery for platinum contents exceeding 95 wt% (approx. 70 at%). In Pd-(Cu,Ni)-P BMGs, increasing m-fragility with higher palladium contents is attributed to the less pronounced bifurcation into Pd-Cu-P and Pd-Ni-P clusters with different coordination shells around the central phosphorus atom (77, 95, 97). These two structural units were suggested to develop MRO that stabilises the supercooled liquid and is responsible for the stronger liquid behaviour (95). Topological-based assumptions that Pd-P and Pt-P liquids have similar structures led to the expectation that the same structural changes would occur in Pt-P liquids (95). In practice, the liquids show substantial differences in SRO and MRO (97, 98).

While icosahedral SRO dominates in Pd-P liquids, Pt-P liquids contain many more trigonal prismatic structural units, leading to the observation of pronounced MRO. This MRO is not observed at high temperatures, unlike the icosahedral SRO in Pd-P liquids, resulting in more pronounced ordering during cooling towards T_g (95, 97). As a result, Pt-P liquids have a much broader distribution of cluster connection schemes, comprising the more flexible two- and four-atom connections, whereas the stiffer three-atom connections prevail in the Pd-P-based liquids (97). The presence of SRO at high temperatures and a relative lack of ordering on the medium length-scale in Pd-P-based liquids lead to a low entropy of fusion compared with Pt-P-based liquids. This explains the low driving force for crystallisation in Pd-P-based liquids (62, 83, 87, 98, 99) as well as the more pronounced sensitivity of plasticity to the cooling rate during glass formation (97, 100, 101). Regardless of the differences when compared with Pt-P-based liquids, the high m-fragility of Pd-P-based liquids does mean that there is still substantial ordering on cooling towards T_g; fragility is dependent on the rate of ordering near T_g, not on the type of ordering (92).

The primary restriction on BMGs for jewellery is hallmarking standards (as practised in the UK) and stamping regulations (US equivalent). These hallmarking standards place a minimum weight fraction on platinum, palladium, gold and silver to guarantee quality (Table II). Of these hallmarks, 18 karat gold, 950Pt, 900Pt, 950Pd and sterling silver are predominantly used for jewellery.

Alongside the requirement for a minimum weight fraction of precious metal, jewellery alloys must also be without skin-sensitising elements such as nickel (42, 45, 102). Consequently, many of the otherwise attractive glass-forming compositions are unsuitable for jewellery. Although studies suggest that the ion release rate of nickel from glassy alloys is well below the legal requirement (45), there is an industry-wide desire to eliminate such elements (103). Nickel-free BMGs are also of substantial interest for dental and biomedical applications (104, 105).

Schroers and Johnson’s discovery of 850Pt BMGs was, therefore, a breakthrough (28). The high GFA, high m-fragility and low T_g of these liquids are desirable for high-definition TPF (41, 71, 106). Low casting temperatures and the glassy structure itself appear to solve many issues in jewellery manufacture (Section 1, Part II (17)), while the unique properties of BMGs can be exploited for scratch-resistance and other property enhancements (Section 2, Part II (17)). Still, it is the high 950Pt and 950Pd hallmarks that are the most desirable for jewellery. Given the inevitably high capital cost that would be associated with the manufacture of BMGs using techniques that are novel and unfamiliar to goldsmiths, BMG jewellery would, at least initially, be expensive. Hence 950Pt and 950Pd hallmark-compliant BMGs would be desirable, if not, essential. It is conceivable that the majority of consumers would justify the cost of these materials by fineness alone. While the 900Pt hallmark is widely used by jewellers and offers a wider ‘compositional space’ for alloy development, it would not command as high a price.

Attempting to achieve such high-fineness hallmarks heavily restricts the extent of alloying, and yet, as discussed above, alloying is necessary to achieve high GFA. The platinum-rich (95 wt% platinum) and palladium-rich BMGs (90 wt% palladium) developed by Demetriou et al. (73, 74) are, therefore, both scientifically and industrially significant. Through microalloying (i.e. minor additions of elements), they were able to achieve high weight fractions of platinum and palladium in fully glassy samples, albeit with relatively poor GFA. While Demetriou et al. are correct that their 950Pt BMG would be suitable for jewellery (73), this risks ignoring practical issues. Techniques such as tilt casting, suitable for industrial jewellery manufacture, cannot achieve the cooling rates achieved by expensive, small-scale, laboratory techniques such as suction casting. In an industrial environment, it will be challenging to cast these alloys into a fully glassy state. Furthermore, jewellery alloys often contain a higher fraction of precious metal than hallmark requirements stipulate. This tolerance ensures that all items produced are hallmark-compliant. Increasing the weight fraction of platinum or palladium any further in these alloys is likely to result in the loss of useful GFA.

The work of Demetriou et al. (73, 74) shows that, with difficulty, 950Pt and high fineness palladium (90 wt%) hallmark-compliant BMGs are achievable, but the best alloys developed so far are still relatively poor bulk glass formers. The necessity to add more than approximately 15 at% metalloids to achieve high GFA means that achieving high fineness is difficult.

Microalloying may be key in achieving better compositions, although its effects are poorly understood. Minor additions can have a significant impact on GFA: adding 0.3 at% silver to Pt_74.7Cu_1.5P₁₈B₄Si_1.5 increases d_c by a factor of two, albeit from a low level (73). Additions of up to 2 at% gold or silver increase GFA in Pd₇₉Cu₆Si₁₀P₅ (86), rationalised using a topological argument (although joint additions yield less impressive results).

The effect of microalloying on properties of BMGs appears specific to the added element (107, 108), suggesting that both chemical and topological effects are important. From a topological viewpoint, microalloying is expected to increase the variety of local atomic configurations (108–110). This may affect local atomic rearrangements (β relaxation) linked to spatial heterogeneity (111) and therefore the properties of BMGs below T_g. Faster β relaxation due to microalloying (109) is expected to significantly affect a wide range of properties, from corrosion and tarnishing behaviour (65, 112) to sub-T_g embrittlement (97, 112–115). If microalloying is to be employed in platinum- and palladium-based systems to permit an increase in the precious-metal weight fraction, a deeper understanding is required of its effect on glass properties and GFA.

Alloys containing phosphorus are challenging to process. Its reactivity means additional processing steps, such as prealloying, are required (11). Pt-P and Pd-P binary eutectics are, therefore, not ideal starting points for glassy jewellery alloys (11).

For platinum-based BMGs, extensive work by Kazemi et al. has identified the Pt-Si-B ternary eutectic as a suitable starting point (11) for alloy development. The eutectic composition lies at a high atomic percentage of platinum, while the weight percentage of platinum is aided by the low atomic masses of silicon and boron. Many other M-Si-B eutectics (M = nickel, cobalt, iron) have high GFA, which is rationalised by topological arguments (67, 116, 117). Prior studies, comparing the GFA of Pd-Si and Pd-Si-B, further support expectations of a high-GFA platinum-based BMG (118). By substituting copper for platinum, and germanium for silicon, at the Pt-Si-B ternary eutectic composition, a Pt-Cu-B-Si-Ge BMG with a d_c of 5 mm was developed (11).

In contrast to phosphorus-containing BMGs, Kazemi et al.’s Pt_49.95Cu_16.65Si_6.4Ge₃B₂₄ alloy is unaffected by fluxing. Prolonged fluxing with B₂O₃ did not reduce the oxygen content (119). While oxygen-scavenging by alloying with scandium was successful in reducing oxygen content, the GFA was not improved (119). Instead, all GFA was lost when more than 2 at% scandium was substituted for copper (119). Scandium and holmium additions were found to sharply increase T_g and T_x, with a concomitant rise in hardness. These additions also led to a rise in T_L, ultimately leading to a loss of GFA (lower T_rg) (119). While oxygen scavenging, minor additions of large-atomic-radius rare-earth metals and prolonged fluxing all improve the GFA of Pt-P (87, 120, 121), their failure to do so for Pt-Si-B suggests that the controlling factors are different. To elucidate these and to optimise alloy compositions, it would be helpful to know the fragility of the liquid: no values have been reported, but the liquid is expected to be relatively fragile.

Apart from Pd-P, only the Pd-Si eutectic has been reported as a suitable starting point for BMGs. Early work by Chen and Turnbull focused on the Pd-Cu-Si ternary eutectic, yielding high-GFA alloys (15). The similarity in weight fraction of palladium between Pd-P and Pd-Si means, as for platinum-based BMGs, that high fineness and high GFA conflict with one another. Many Pd-Si-based glass formers have low weight fractions of palladium and are therefore not suitable for jewellery. They have, however, attracted much interest for their use as biomedical and dental materials due to their high plasticity, absence of skin-sensitising elements and excellent corrosion resistance (104, 105).

Although high fineness is still challenging, the Pd-Si binary eutectic serves as an excellent starting point for BMGs. Pd-Si binary alloys are remarkably good glass formers; the glasses exhibit high activation energies for crystallisation (122), resulting in high thermal stability. However, only after a careful choice of elements with low atomic mass and suitable atomic radii is the suggestion of a 950Pd hallmark-compliant BMG plausible.

Reports of Pd₇₉Au_1.5Ag₃Si_16.5 (d_c = 3 mm) (105) and Pd₇₅Si₁₅Ag₃Cu₇ (d_c = 10 mm) (123) BMGs support these claims. These are two of the highest fineness phosphorus-free palladium-based BMGs reported so far, with weight fractions around 90 wt% palladium before further alloying. Pd₇₅Si₁₅Ag₃Cu₇ is also readily formed using TPF and does not require B₂O₃ fluxing, making it easier to manufacture (123). While a phosphorus-free BMG would be desirable, small additions of phosphorus can improve GFA and permit an increased weight fraction of palladium (124).

Further research on 950Pt and 950Pd hallmark-compliant BMGs with a GFA suitable for industrial manufacture is challenging. The inevitable high m-fragility of these alloys means that high cooling rates and carefully optimised processing are likely to be requirements. Research has mostly focused on Pt-(Cu,Ni)-P and Pd-(Cu,Ni)-P based BMGs (16, 20, 41, 42, 73, 74, 84, 89, 90, 104, 125) with some interest shown in Pt-Si-B and Pd-Cu-Si based BMGs (85, 86, 88, 105, 123, 124).

Further optimisation through careful alloying and microalloying may help to identify additional 950Pt BMGs and novel 950Pd BMGs with sufficient GFA. While other systems such as Pd-Ni-S BMGs or novel high-entropy BMGs exist, they have low palladium contents and so do not seem suitable for jewellery (5, 126). Li et al. suggest that there may exist up to 10⁶ BMG compositions (127), so the search for new glass-forming compositions should continue. New combinatorial techniques to assess ΔT_x and T_rg (128–130), and computational techniques (131–133), allow more efficient sampling of compositional space for undiscovered high GFA compositions. A new search with these techniques may yield results that cannot readily be predicted and rationalised using empirical rules and experimentally determined phase diagrams.