TECNALIA developed a two-stage leaching and reduction process using deep eutectic solvents (DES), comprising choline chloride and oxalic acid (Figure 2). The solvent is first used to leach the pgms from the solid feedstock at a temperature <90ºC, followed by a reduction of the pgms to their metallic form, by heating the solution to >100ºC. Centrifugation was employed at this laboratory scale based on equipment availability, separately both for the removal of the depleted autocatalysts substrate and the pgm product. Recovered DESs can be recycled for the next leaching batch.

Fig. 2

The solvents used are readily prepared from renewable, non-toxic and naturally occurring chemicals when compared to traditional hydrometallurgical processes that employ strong acids (15, 31). Another advantage is that it is a one-pot leaching and reduction process where no additional reducing agent is required. An increase in the pgm purity is achieved from the selective reduction process with a highly pure (90–95%) pgm solid generated, a significant improvement from the low content feedstock. Therefore, it is a simple process for the concentration of pgms, with reduced costs and wastes generated, compared with the cementation process (31–33) and with a very low energy consumption compared to existing pyrometallurgical processes (15). Further separation steps are needed to separate the three pgms from each other. The DES must be kept close to the leaching operation temperature to avoid handling issues arising from DES solidification.

An advanced leaching process using microwave (MW) technology has been developed by VITO using a laboratory MW digester (flexiWAVE, Milestone Srl, Italy). The feedstock is contacted with 6 M HCl and H₂O₂, as an oxidising agent, before being heated using MW radiation (2.45 GHz) to the reaction temperature of 150ºC. MW heating allows short leaching times, consisting of 15 min heating with 10 min dwell time at the reaction temperature (34). The resultant leachate, containing the pgms, is filtered, to remove the undissolved catalyst substrate, leaving an aqueous solution containing the pgms in a chloride matrix.

The MW heating promotes fast, homogeneous volumetric heating enhancing leaching reproducibility and efficiency, with a positive impact on the energy efficiency. Optimisation of the leaching process has resulted in a significant reduction in the HCl acid concentration (i.e. 6 M HCl) required compared to previous results reported in literature (i.e. 12 M HCl) (34). The necessity of the H₂O₂ addition is dependent on the feedstock (Table I, data provided in presence of 10 v/v% H₂O₂ (31%)). In some circumstances its addition has been shown to have a significant impact on the pgm leachability. Through modelling and analyses of the reactor head-space gas, it was shown that H₂O₂ addition increased the formation of Cl₂ and H₂ gas, which must be considered in the process safety assessment.

Solvometallurgy is an alternative branch of metallurgy that uses non-aqueous solutions (35). KU Leuven investigated the oxidative dissolution of pgms from the feedstock using two different solvometallurgical approaches. In the first, an innovative method for the selective leaching of palladium from the feedstock was developed, avoiding the co-extraction of platinum and rhodium whose recovery can be achieved from the palladium depleted feedstock by varying the process conditions. Dilute and concentrated solutions of FeCl₃ in acetonitrile are used for the selective dissolution of palladium and the dissolution of platinum and rhodium, respectively. This solvometallurgical approach provides a preconcentration of the pgms after which further purification is needed to separate platinum and rhodium. In the second approach, highly concentrated solutions of AlCl₃·6H₂O and Al(NO₃)₃·9H₂O were used to dissolve palladium selectively from spent autocatalysts; 95% palladium was leached in only 15 min at 80ºC (36).

Ionic liquids (IL) are perceived as ‘green solvents’, mainly due to their low volatility, which signifies a reduction in the degree of negative environmental impact and health hazards, low toxicity and non-flammability (37). In metal extraction applications, they have provided dramatically higher extraction efficiencies than commonly used solvents. They can act either as solvents in the presence of an extracting agent or as selective extractants, since via modifications in their anion or cation their selectivity towards a certain metal can be tuned (38, 39).

VUT explored the properties of ILs for both leaching and separation processes (for separation see Section 2.5). Leaching of the pgms from the feedstock was performed using hydrophilic and low cost choline-based ILs. The choline-based IL leaching process was selected as the optimum due to its high extraction efficiency and selectivity. The process operates at mild conditions in the presence of an oxidising agent: <100ºC for 4 h in a sealed vessel with continuous stirring (Figure 3).

Fig. 3

This process leaches the pgms, but further separation steps are needed to separate the three pgms from each other. Depending on the feedstock, the loaded IL can be re-used for subsequent leaching batches until the IL capacity is reached. Of note is the mild process conditions used.

The major drawback is the high IL viscosity, making implementation on an industrial scale challenging. Nevertheless, dilution of the IL with water is feasible without compromising the pgm extraction efficiencies. Furthermore, larger scale system design and operation must ensure the violent exothermic reaction between the feedstock and oxidising agent is managed safely.

Due to the increased interest in ILs, a new concept surfaced: IL immobilisation on solid support materials, which is the deposition of a fine IL layer on a solid surface. Use of supported IL phases (SILPs) is an ideal strategy in order to make use of the benefits of IL and simultaneously avoid the inherent complications of IL-based separations such as mass-transport limitations and excessive usage of the IL (40).

VUT exploited the properties of hydrophobic ILs in both liquid and solid extraction processes; this was employed in conjunction with the VUT ionometallurgical leaching process (Section 2.4). In the liquid-liquid extraction process, a solution comprising 50 wt% phosphonium-based IL in n-heptane was used to extract palladium, platinum and rhodium from the loaded pgm hydrophilic leachate (generated from ionometallurgical leaching) to the hydrophobic IL phase. Quantitative extraction is obtained after continuous stirring for 2 h at room temperature. The pgm-free hydrophilic IL phase can be recycled to perform the next batch of IL leaching. No significant performance loss of the hydrophilic ILs was observed between the different cycles; only five cycles have been tested to date. This is particularly attractive to minimise environmental impact and operating costs.

The alternative approach, solid extraction, relied on SILPs using hydrophobic phosphonium-based ILs (Figure 4). The pgms are adsorbed onto the solid enabling their one-step separation from the main impurities of aluminium, iron and cerium due to their low retention on the solid material.

Fig. 4

The stripping is a two-step process: (a) an acidified thiourea solution strips the impurities retained on the SILP; (b) a more concentrated acidified thiourea solution strips palladium and platinum (rhodium is retained on the solid).

The use of SILPs allows fast and simple separation of the pgms from other metals with reduced chemical reagent consumption compared to liquid-based separations. There is the possibility to re-use the solid material for further separations without any loss in its retention and separation performance, to date one recyclability experiment has been conducted. An advantage of SILPs compared to most commercially available resins is no pre-equilibration is required; this could have significant impact on cycle times and effluent generation if re-equilibrium is necessary, as part of the load-elution cycle. A drawback of the process is the retention of rhodium alongside chromium on the SILP and the removal requires a suitable stripping agent not yet identified.

A split-anion solvent extraction process has been developed by KU Leuven for the separation of an aqueous mixture of pgms into their individual elements (Figure 5) (41). Generally, the split-anion extraction relates to the solvent extraction process where different anions are present in the aqueous and organic phases and the distribution of the IL anions strongly favours the IL phase. The pgms are extracted from the chloride leaching solution using the iodide form of the quaternary ammonium IL Aliquat^® 336, [A336][I], dissolved in p-cymene. The iodide anions, which have a strong affinity for the organic phase, coordinate with pgms to form stable iodo-complexes that can be extracted to the ionic liquid phase (42, 43). The split-anion extraction allows not only efficient extraction of pgms without changing from a traditional chloride feed solution, but also the selective recovery of the extracted metal complexes from the loaded organic phases. The organic is scrubbed and stripped of its pgms, as detailed in Figure 5.

Fig. 5

The ionic liquid-based split-anion extraction process is simple, selective and effective for the sustainable separation of pgms, using only one ionic liquid [A336][I] as the extractant, which can be regenerated for consecutive extraction-stripping cycles. The high viscosity of [A336][I] is a drawback, which has shown, during its pilot scale application in mixer-settlers, to slow the mass transfer. Some measures have been identified to reduce the IL viscosity, such as the use of water-saturated ionic liquid or use of green diluents (i.e. p-cymene), but the measures can only partially resolve the problem. To the best of our knowledge this is the first time that a process based on IL for the separation of pgms is tested in continuous mode using real pregnant leach solutions as feed. Other separations using IL have been developed and show good performance but have been only tested with synthetic solutions and not in continuous mode (44, 45).

Sorbents are an established method to selectively recover palladium and other noble metals for acidic aqueous streams such as leachates (16, 46–48). Many materials have been developed by different groups and their performance tested in powder form. However, to be applied in a continuous way, the shape has to be optimised to avoid clogging and pressure build-up in a column or the material has to be modified for easy recovery afterwards with for example a magnetic core (49, 50). Therefore, VITO has developed solid sorbents by first selecting or forming a suitable solid backbone before grafting active organic scavenging groups onto the support using a green aqueous synthesis method. Two types of regenerable three-dimensional (3D) structured metal oxide supports (Figure 6) have been investigated: monodisperse microspheres and 3D printed monoliths. These types of supports have the advantage of reducing mass diffusion limitations, optimising packing density and decreasing pressure drop while allowing fast adsorption and desorption cycling times.

Fig. 6

The functionalised microspheres show good selectivity for palladium over other pgms and impurities giving a method to remove palladium from acidic solutions. The adsorbed palladium can be easily recovered by stripping with concentrated HCl acid resulting in a concentrated acidic solution of palladium. The sorbents have a palladium capacity of 0.33 mmol g^–1 from acidic solutions (pH 2). When stripped with 3 M HCl, a seven-fold increase in the palladium concentration between the feed (100 mg l^–1 palladium) and stripping solution was observed. No significant performance loss of the sorbents was observed between the different cycles; only seven cycles have been tested to date.

The developed material is a hybrid (ceramic-organic) adsorbent selective to palladium over the rest of the pgms with good hydrolytic stability that can be regenerated and reused. If a degradation of the organic scavenging groups and hence, decreased performance is observed over a higher number of cycles, the metal oxide support can be recovered and refunctionalised. The biggest drawback is the limited sorbent capacity, but more developments to improve this are ongoing by increasing the specific surface of the supports and optimising the grafting conditions.

SINTEF investigated extracting pgms from the feedstock using pyrometallurgy, employing copper as the pgm collector, followed by a molten salt electrolysis process (Figure 7).

Fig. 7

The pyrometallurgical process was investigated at two laboratory scales (10 g and 5 kg). The pgm recovery rates, in the alloy phase, were close to 100%, and the copper-collector recovery rates were in the range 82–100%, when using the optimised parameters: 10 wt% copper-collector, 10–15 wt% calcium oxide, 1600–1650ºC and 1–1.5 h holding time. This pyrometallurgy step pre-concentrated the pgms by ~10 times from feedstock to the generated metallic phase.

In the electrolysis step, the copper-pgm alloy is used as anode in an electrorefining cell with the eutectic LiCl-KCl as electrolyte at 450ºC. The experiments demonstrated the selective extraction of the metal phase (copper), which was recovered at the cathode with a current efficiency of ca. 70%. Under these conditions, the pgms (and other impurities) remain in the anode residue giving a solid with >99.9% purity. Further separation steps are needed to separate the three pgms from each other.

One significant advantage of pyrometallurgical over hydrometallurgical processes is the lower reagent use in relation to the feedstock pgm content (kilograms of reagent per kilogram of pgm) (51). As such, it provides attractive conditions for preconcentrating pgms from very dilute wastes, such as the PLATIRUS feedstocks; the optimised conditions of the SINTEF process showcase that. High recovery rates of the copper collector at the electrolysis cathode have been achieved and it can be recycled for the next pyrometallurgical step.

The electrolysis process allows the extraction of copper from the pgm-containing copper anode in a molten salt electrolyte with better selectivity and kinetics as well as lower energy consumption than in state-of-art copper-refining processes using aqueous solutions (52). Cu(I) species are stable in the molten salt electrolyte, thus the voltage (and energy) needed in the electrorefining process is lower than in an analogous aqueous solution process where Cu(II) are the solely stable species. In general, the kinetics of the electrode charge-transfer reaction in molten salts are considered faster due to the high operational temperature (53).

The challenges of operating the SINTEF pyrometallurgical step are the same as those found in industry. The separation of pgm microparticles from the molten slag phase is impacted by the slag viscosity and metal-to-slag interfacial tension and in turn affect the pgm extraction efficiencies. The energy consumption was ca. 5.5 kWh kg^–1 copper recovered in the pyrometallurgical step and ca. 7 kWh kg^–1 pgm recovered in the electrolysis step (equal to 0.3 kWh kg^–1 copper refined) under optimised experimental conditions. Energy consumption can be challenging to extrapolate from laboratory to industrial scale. Due to the energy intensiveness of pyrometallurgical processes, energy consumption and associated cost must be evaluated at larger scale.

Molten chloride mixtures can be used as a reaction media in the chlorination of oxide mixtures, ores or industrial byproducts. The dissolution reaction generates chloride compounds at much lower temperatures (ca. 450ºC) than those needed in solid-gas chlorination reactions (ca. 1000ºC). This is due to significant solvation effects of the dissolved metal cation with the chloride ions of the molten chloride media.

SINTEF investigated the selective recovery of the pgms using molten salts and chlorine gas as an oxidiser and chlorination agent, followed by an electrolysis process (Figure 8). LiCl-KCl eutectic mixture was chosen as the best candidate. The feedstock was fed into the reactor at 450ºC without any pretreatment or up-concentration steps, resulting in a single step pgm extraction process. Silicon, magnesium and aluminium were not dissolved thus remain as a solid sludge at the bottom of the reactor. The analysis of the residue shows that, after 3 h, the total chlorination and therefore dissolution of the pgms is close to 50%. The dissolved pgms are recovered in an electrolysis process as a metallic-pgm alloy at the cathode resulting in chlorine evolution at the anode. Only 80% of the dissolved pgms can be accounted for in the molten chloride as pgm-chlorocomplexes; it is believed unaccounted pgm mass corresponds to formed volatile pgm-chlorocomplexes that could be recovered from the off-gas system by condensation; estimates calculated from a mass balance suggested a feed volatilisation of ca. 10% palladium, 10% platinum and 20% rhodium.

Fig. 8

This chlorination process presents clear advantages in terms of rate, conversion and selectivity when compared with traditional gas-solid reaction systems. In the latter, the rate and conversion are limited by the contact of the gaseous chlorinating agent and the material to be chlorinated, and the reaction occurs non-selectively as, at the much higher temperature, chlorination of all other elements contained in the material occurs, not only pgms.

Though faster than gas-solid reactions, the kinetics are still slow when using chlorine gas and the use of other gaseous chlorination agents, such as HCl, should be tested. Further optimisation of the process is required to achieve competitive pgm recovery rates, including recovery of the volatile pgm-chlorocomplexes.

GDEx is defined as the reactive precipitation between metal precursors in solution and intermediates from the reduction of gases at a gas-diffusion electrode. When the gas is air or O₂, the O₂ reduction reaction leads to hydroxyl ions and hydrogen peroxide being formed in the pgm solution, which react, forming oxides or hydroxides (Figure 9). Alternatively, the process can run with other gases. The GDEx process is enabled by VITOCoRE^® multi-layered gas-diffusion electrodes (54).

Fig. 9

GDEx uses the cleanest possible reagent, the electron, and it is highly versatile, as it can be used to recover many different metals. The process uses an inexpensive reactor. As it operates in a flow-cell configuration, it is easily up-scalable by stacking multiple individual cells, without a reduction in performance. The process is highly reproducible, involves mild operation conditions (room temperature and atmospheric pressure). The process has a low energy consumption, for example ~2-6 kWh kg^–1 of materials recovered, when compared with electrowinning platinum from chloride media reported at 21 kWh kg^–1 platinum (55). Finally, it is efficient, i.e. 50% current efficiency for the formation of the reactive intermediates that fully react with the metal precursors to achieve the targeted recovery.

Notably, the recovery of dilute metals and simultaneous synthesis of nanostructures with GDEx is fast, with rates approaching ~3–15 kg per day, using a single, inexpensive, electrochemical reactor under flow regime.

The best results were achieved with dilute metal concentrations, and significant optimisation is required for solutions with metal concentrations above 10 g l^–1. Especially with high metal concentrations, it is expected that, after a period of operation, the electrode would become clogged and require an acid treatment to regenerate the electrode porosity. The recovery of the precipitated materials is impacted by the unoptimised downstream separation and drying processes. Some reagents, such as sulfur-based compounds, are known to interfere with the process under defined processing conditions, coprecipitating with the pgms.

A laboratory scale MW system (flexiWAVE) equipped with a spinning carousel holding 15 pressure-sealed Teflon-lined reactors was used to process ~1.3 kg autocatalyst (Figure 13).

Fig. 13

The leaching temperature is measured by a thermowell contained optic fibre, placed in one of the 15 reactors. Each of the 15 reactors was loaded with 5 g of the feedstock and 50 ml 6 M HCl solution. The MW-assisted reaction took place at 150ºC for 10 min, with a heating time to the set temperature of 15 min. Subsequently, the reactor was cooled, opened and the leachate was vacuum filtered. It is noteworthy that this leaching process did not require addition of H₂O₂ as an oxidation agent and thus the formation of hazardous head space gas mixtures (containing H₂ and Cl₂ gas) was avoided. During vacuum filtration, the leach residues were washed with 6 M HCl at room temperature. Both leachate and washing liquids were collected.

Overall, ~1.3 kg of material was leached by performing 18 MW-leaching runs. Three batches of material with different grain sizes and total masses were processed, hence three leachates (L1, L2, L3) and washing waters (W1, W2, W3) samples were obtained (Table III).

The leachate containing L1, L2 and L3 (Table III), as prepared by VITO, was processed using the solvent extraction flowsheet developed at KU Leuven (Figure 5). Due to the low pgm concentration of the leaching residue washings, these solutions were not processed. A continuous solvent extraction demonstration using multi-stage mixer-settlers was undertaken to process the leachate (Figure 14).

Fig. 14

First, Pd(II) and Pt(IV) were quantitatively extracted in two countercurrent stages with [A336][I] in p-cymene at O/A = 1/3, leaving Rh(III) in the raffinate. The impurities (mainly aluminium, barium, cerium, iron and tin) in the loaded organic phase were removed with NaCl solution. Then selective stripping of palladium was achieved by equilibrating the scrubbed organic phase with aqueous ammonia in NaCl solution in four-stage mixer-settlers at O/A = 3/1. Followed by the recovery of platinum from the palladium-free loaded organic phase was performed using acidic thiourea solution [CS(NH₂)₂/HCl] in four countercurrent stages (O/A = 2/1). After being washed with diluted HCl, the IL requires regeneration by contact with KI to replace the chloride for an iodide anion reforming [A336] [I].

The work at KU Leuven demonstrated that continuous solvent extraction is feasible for not only the preparation of [A336][I] but also the selective recovery of individual pgms, with recoveries of 81 ± 3% Pd(II), 62 ± 2% Pt(IV) and 54 ± 5% Rh(III), from the feedstock leachate. The mass loss was mainly as a result of the samples taken for analysis and the mixer-settlers start-up and finish steps. These losses are not representative of industrial scale operation in which mixer settlers would operate continuously.

The rhodium raffinate and platinum strip were further processed using conventional chemical transformations to convert the pgms to the nitrate form. The palladium strip was sent for further processing by the GDEx process.

GDEx attained a recovery of ~70 ± 1% of palladium from the KU Leuven strip sample (Figure 15). Palladium was recovered with a purity of 91–93%, with platinum, rhodium and aluminium as the major impurities, and additional minor impurities of barium, iron, magnesium, cobalt and copper. Only two batches were processed in comparison to the eighteen leaching batches performed thanks to the up concentration in the solvent extraction process. The selectivity of the recovery can be optimised by further investigation into the effect of the different GDEx operational variables, such as the influent concentration, hydraulic retention time, applied potential or current.

Fig. 15

The palladium sponge generated from the GDEx technology was further processed using conventional chemical transformations to convert it to palladium nitrate.

The platinum group metal iridium is perhaps the most puzzling metal on Earth due to its property of being cleavable and a plastic solid simultaneously (1). This refractory face-centred cubic (fcc) metal (T_melt = 2446°C) serves as the structural material for applications under extremely hard conditions (2, 3) such as containers for fuel sources in radioisotope generators for deep space missions (4), or crucibles for growing oxide crystals for power lasers (5). Industrial technology for refining and processing iridium, based on traditional chemical refining methods (2, 6, 7), has been developed over the past 60 years (8–10). Based on these achievements, it has been shown that polycrystalline iridium exhibits limited plasticity due to intergranular fracture at room temperature, but its plasticity increases considerably under elevated temperatures (9–12). The segregation of non-metallic impurities on the grain boundaries was considered the cause of poor workability of polycrystalline iridium (12). This type of deformation behaviour agrees with empirical knowledge on deformation and fracture of metals (13, 14). On the other hand, single crystalline iridium behaved unusually: it cleaved under tension after considerable elongation (13, 15), but never failed under compression (16–18). At room temperature the fracture mode of iridium single crystals was attested as BTF (1, 15), while brittle intergranular fracture (BIF) was the fracture mode for polycrystalline iridium (11, 19). Analysis of the causes of cleavage in iridium has shown that it satisfies some empirical cleavage criteria (18–20) due to features in the elastic moduli in comparison with other fcc metals (18, 21). This fact leads to the conclusion that the inclination to cleavage is an intrinsic property of iridium, whereas impurities only reinforce it (18–20, 22, 23). However, the analysis of interatomic bonding in iridium has shown that BIF may also be considered as the intrinsic fracture mode of polycrystalline iridium (24).

The pyrometallurgical scheme for the refining of iridium, including: (a) oxidation induction melting; (b) electron beam melting; and (c) growing massive single crystalline workpieces by electron beam, became an alternative technology to manufacture ‘plastic’ iridium (25–27). Pyrometallurgical iridium demonstrated considerable plasticity prior to failure under tension in both the single crystalline state (28) and the polycrystalline state (29), while its elastic properties were the same as findings obtained earlier (30). It was confirmed that the intrinsic fracture mode of this ‘plastic’ iridium is BTF, while BIF is induced by harmful non-metallic impurities such as carbon and oxygen (31, 32). Indeed, the portion of BTF on the fracture surface of plastic iridium is considerably higher than BIF (33, 34). Deformation mechanisms and, hence, behaviour of pyrometallurgical iridium were the same as a normal fcc metal excepting the special fracture mode (35–38). Recent studies of deformation and fracture behaviour of iridium have shown that new participants achieved the technological level that allows ‘plastic’ iridium to be manufactured (39, 40), including iridium single crystals (41). Also, the old problem concerning the intrinsic fracture mode of polycrystalline iridium or the competition between BTF and BIF in iridium remains (42). Therefore, in the present paper, the deformation and fracture behaviour of iridium wires under tension at room temperature are considered in light of the discussion on this problem.

Pyrometallurgical iridium was used in this work. It was high purity metal, free of non-metallic contaminants such as carbon and oxygen. The refining procedure and, hence, impurities content were the same as the metal used in earlier work by our group: non-metallic elements <0.1 ppm; tungsten, molybdenum, niobium, iron, zirconium, copper, gadolinium, yttrium, gallium, nickel, palladium, zinc, magnesium, calcium –0.1–1 ppm; platinum, rhodium ~10 ppm (26, 27, 29, 38). Experience has shown that pyrometallurgical refining could be limited by the first and second procedures without loss of quality of the metal. Therefore, the operation of the growth of the massive single-crystalline iridium workpieces was not carried out in this work. The mechanical treatment of the pyrometallurgical iridium included: (a) forging the ingot into sheet at 1500–2000°C in air; and (b) rolling the sheet at ~800°C in air. The resulting metal could be processed like platinum. The cold drawing iridium wire, whose diameter varied from 2.7 mm to 0.5 mm, was prepared from this plastic metal. No long-term recrystallisation annealing of this wire was carried out because this procedure leads to embrittlement and failure of iridium wire due to BIF. Tensile testing was carried out with the help of an Autograph AG-X 50N tensile/compression tester (Shimadzu Corporation, Japan) (traverse rate of 1 mm min^–1) at room temperature. The lengths of working parts of iridium wire samples were 100 mm. The structure of the samples before and after testing was examined by conventional X-ray diffraction (XRD) technique on the D8 Advance diffractometer (Bruker Corporation, USA) with copper kα irradiation. Back surfaces of each sample before and after testing were documented on a light metallographic microscope. The fracture surfaces of samples were studied on the scanning electron microscope JSM-6390 (JEOL Ltd, Japan).

The first set of iridium samples consisted of 10 pieces taken from a commercial parcel of cold drawing thin iridium wire produced by UralInTech (Russia) having a diameter of 0.5 mm. The microstructure of this wire had a strongly deformed lamellar morphology, where the grains of the polycrystalline matrix practically disappeared (Figure 1). The main feature of this lamellar structure is the narrow highly elongated grains collected in a bunch like a rope. As a result, deformation tracks, such as slip bands or twin lamellae, could not be revealed on the surfaces of the samples after deformation. An XRD spectrum taken from the cold drawing iridium wire prior to testing is shown in Figure 2. There are two high narrow peaks ((200) and (220)) in the middle angles of the spectrum taken from the sample. No visible changes in the spectrum were revealed after tensile testing of the sample. It may be concluded that a stable drawing texture is formed in the iridium wire in comparison with an annealed polycrystalline sheet, which does not depend on further tensile deformation.

Fig. 1.

Fig. 2.

The second set of iridium samples consisted of 10 cold drawing wires with a diameter of 2.7 mm taken from the workpiece that was used to manufacture the thin plastic iridium wire. The Vickers microhardness of these samples in the undeformed state was about 7 GPa. The third set of iridium samples contained 10 cold drawing wires with a diameter of 2 mm. In contrast with the second set, these samples were annealed at 1000–1200°C for 20 min in a low vacuum and, as a result, their Vickers microhardness dropped up to 5 GPa. This operation is also used in the technological process for the manufacture of plastic iridium wire.

The stress-strain curves of the cold drawing iridium wires are shown in Figure 3 and some of their mechanical characteristics are collected in Table I. The back surfaces of the deformed samples are shown in Figure 4, while their fracture surfaces are given in Figure 5. It is clearly visible that the deformation behaviour of the samples from the first set (Figure 3, curve A) is similar to the behaviour of annealed copper wire (Figure 3, curve B). The long stage of plastic flow takes place after the short stage of material strengthening (Figure 3, curves A and B, respectively). Indeed, the total elongation of both materials may be estimated as considerable for a polycrystalline wire sample (30% for iridium and 43% for copper). In addition, there is a clearly visible advanced necking region on the back surfaces of the deformed samples (thinning of 20% for iridium and 55% for copper) (Figures 4(a) and 4(b) and Table I). However, in contrast with copper, iridium exhibits much higher yield stress and ultimate tensile strength (Table I). In spite of the features that are inherent to the ductile deformation behaviour, the fracture mode of the iridium samples from the first set is attested as BTF in the strongly deformed lamellar structure (Figure 5(a)). The same findings were obtained for cold drawing iridium wire with a diameter of 0.3 mm in the temperature range 20–800°C in (29).

Fig. 3.

Fig. 4.

Fig. 5.

The deformation behaviour of thick cold drawing iridium wire (Figure 3, curve C) can be attested as brittle: its stress-strain curve has an almost rectilinear profile, the yield stress is similar to the ultimate tensile strength, while the deformation prior to failure is small in comparison with the previous case. No necking was observed on the back surfaces of the deformed cold drawing thick iridium wires (Figure 4(c)). The fracture mode of the samples agrees with their brittle behaviour, it is BTF (Figure 5(b)). The short-term vacuum annealing of the thick cold drawing iridium wire at a temperature close to the point of recrystallisation of iridium leads to a change in mechanical behaviour from brittle to ductile. Indeed, the behaviour of the stress-strain curve becomes similar to annealed copper (Figure 3, curve D) when after a short stage of strengthening follows the plastic flow stage, while the yield stress and the tensile strength drop considerably (Table I). However, its deformation prior to failure is very small (Table I) for a plastic material and the neck is absent in the deformed samples (Figure 4(d)). The fracture mode does not change from brittle to ductile: it is attested as a mixture of BTF and BIF (Figure 5(c)).

It was shown that the deformation behaviour of cold drawing iridium wire under tension at room temperature depends on its structural state. Wire with grains of 50–100 μm behaves as a brittle material and exhibits BTF as the fracture mode. Vacuum annealing at 1000–1200°C causes a drop of yield stress of the iridium wire, but does not lead to significant increase of plasticity, while its fracture mode continues to be brittle. On the other hand, thin cold drawing iridium wire having a lamellar structure demonstrates considerable elongation prior to failure and clearly visible necking, despite BTF as its fracture mode. This finding gives the basis for the conclusion that such lamellar structure is the correct morphology for plastic polycrystalline iridium. It is important to note that this morphology is formed in the iridium workpiece under the cold drawing process, while a few short terms annealing at 1000–1200°C are included in the procedure after some rolling passes (29).

Earlier, it was shown that BIF is the impurities induced fracture mode of high purity polycrystalline iridium (31, 32). Indeed such non-metallic elements as carbon and oxygen contained in a low vacuum (10^–2 MPa) induce grain boundaries brittleness, but the kinetics of the process depends on the working temperature and its duration (31). For example, under annealing of 20 min at 1200°C, the portion of BIF on the fracture surface is considerably less than the portion of BTF, while after 24 h annealing BIF covers the whole fracture surface. It means that the regime of annealing of the cold drawing iridium wire used in this work is optimal because the hardness and the yield stress of the iridium workpiece are decreased, but the cohesion strength of grain boundaries does not drop.

The low plasticity of the thick cold drawing iridium wire, which was strongly hardened during preliminary processing, may be explained by the supposition that its resource of plasticity is finally exhausted under tension as it takes place in iridium single crystals under the same experimental conditions (26, 37). As a result, the cleavage crack can appear on any dangerous macroscopic surface defect and, hence, such wire is prone to separation by the brittle route without necking. Following this logic, the thin cold drawing iridium wire should behave the same way; however, it exhibits ductile mechanical behaviour, except its fracture mode. One cause of this puzzling effect may be the special configuration of the defect structure of iridium, whose feature on the microscopic level is a lamellar morphology. Indeed, high purity polycrystalline iridium is able to undergo severe deformation under high-pressure torsion at room temperature when the nanocrystalline structure is forming in the material (38). It is puzzling, but in this case, the surface defects play the role of the initiation of cleavage in the neck region only (29). Indeed, iridium meets some empirical cleavage criteria (18–20). However, this effect should be considered as an artefact because in contrast with other cleavable solids iridium is a plastic material in both the single crystalline and polycrystalline states and its inclination to cleavage depends on the structural state.

The lamellar structure that forms in iridium wire during the cold drawing process provides the excellent mechanical properties of polycrystalline iridium under tension: it behaves like a ductile fcc-metal excepting the brittle fracture mode. It was shown that the inclination of iridium to cleavage depends on its structural state.

The lattice dynamical study of metallic crystals is an interesting field of research. Platinum group metals are highly valuable transition metals which have many useful properties. The electronic structure of the platinum metals is of impressive theoretical and practical importance. Dependable thermodynamic information is expected to give the crude material from which lattice dynamics, electronic conveyances and energy states can be deduced by genuine understudies of the solid state. In the present manuscript the author has used VTBFSM for theoretical calculation. The pioneering work of Kellerman (1) for ionic interactions in the alkali halides has attracted considerable attention theoretically as well as experimentally. Löwdin’s (2, 3) and Lundqvist’s (4–6) theory for ionic solids leads to the first important term as many body force which includes the three-body component. The Heitler-London theory and the free-electron approximations will employ the combined effects of VWI and TBI in RSM (7). The effects of VWI and TBI in the framework of ion polarisable RSM (IPRSM) are effective up to the second neighbour with short-range, VWI and TBI interaction. The experimental investigation for the phonon dispersion curve of platinum has been done with coherent inelastic neutron scattering, variation in Debye temperature and Raman spectra (8–10). The elastic constants and dielectric constants (11), physical and natural properties of platinum have been elucidated by expedient of theoretical models (12–16), which has also successfully described their interesting properties. After the failure of the Kellerman rigid-ion model (RIM) then Karo and Hardy (17) used a deformation dipole model, Woods et al. (7) and Dick and Overhauser (18) used a RSM to report lattice properties of alkali halides. The other most prominent model was also proposed by some researchers, among them Schröder’s (19) breathing shell model, Basu and Sengupta’s (20) deformable shell model and the three-body force shell model of Verma and Singh (21–23) and Singh et al. (24) for such halides. In consideration of the effect of VWI, reported by Upadhyaya et al. (25), excellent results have been procured between experiment and theory for ionic halides and semiconductors. The betterment of the present model VTBFSM over others can be realised from the fact that in the present model relatively fewer numbers of parameters have been able to interpret numerous and largely divergent physical properties of materials. This has to motivate the author to incorporate this model in the present study.

In the present paper the parameters including (C₁₁, C₁₂ and C₄₄), polarisabilities (α₁, α₂), and lattice constant by (26) have been theoretically calculated for platinum and given in Table I. By solving Equations (i) and (v) we can obtain the phonon spectra in the first Brillouin zone divided into evenly spaced miniature cells. The theoretical results were obtained by VTBFSM. We have used the computed vibration spectra to study the specific heat and infrared (IR)/Raman spectra in the present paper. The DOS have been obtained by computing the DOS of the frequencies from the knowledge of lattice vibrational frequency spectra. The values of frequencies are compatible with theoretical and experimental peaks and Cauchy-discrepancy for lattice dynamics of platinum.

In the Brillouin zone surface the calculated phonon dispersion curves for platinum are shown in Figure 1 by using first principles along two high symmetry directions (qqq) Г-X-Г. The parallel vibrational modes show real dispersion with a maximum cleave. The upper branch consists of longitudinal modes, while the lower one is the shear‐horizontal mode, along both the Г-X-Г directions. We find that the surface modes for clean platinum (qqq) undergo a few changes in L-T modes on the clean surface, near the zone boundaries and along the X-direction, are replaced only in the dispersion curves. This is because the zone boundaries are moderate and the next two surface modes are strengthened. The experimental reported results for dispersion relation are shown in Figure 1(b) (27). On comparing with the experimental result i.e. Figure 1(a) with Figure 1(b) good agreement can be observed.

Fig. 1.

The DOS vs. frequency curve for platinum theoretically calculated in the energy range from approximately –7.0 eV to 0.5 eV in bulk is shown as a solid line in Figure 2. The bulk DOS exhibits three main peaks that are accurately produced. The small observed differences are related to the shape of the main peaks. The differences between the calculated and observed values were found and discussed in Figure 2. There are very important features obtained from the DOS vs. frequency curve. The information about the surface and resonance states was found through these differences. Below the Fermi level E_F, resonance-states are expected to be obtained, mainly because these energies represent the continuum and few energy gaps exist at these energy values. The experimental reported DOS curve (28) may be compared with the present model i.e. Figure 2(a). Sharp peaks can be seen in Figure 2(a) while in Figure 2(b) the distortion in peaks can be seen which justifies the superiority of the present theoretical study.

Fig. 2.

The specific heat and Debye temperature Θ_D have been calculated as a function of temperature T from the lattice frequency spectra as shown in Figure 3. Debye temperature Θ_D is calculated from different frequency values. The specific heat value of platinum has been measured at extended temperature (0 K to 300 K). The calculated result is in reasonable agreement at moderate temperature and at very low temperature. The comparison can be seen through experimental results (27).

Fig. 3.

The varying investigated properties are distinctly shown in the present study by successful use of VTBFSM, which has provided the complete lattice description of platinum. It agrees well with the test of anisotropy factor A = 2C₄₄/(C₁₁–C₁₂) > unity and towards the high frequency end the higher peak is found. The determined model parameters in Table I were used to solve the secular equation for specified values of wave vectors in the first Brillouin zone, which is split up in an evenly spaced sample of (1000) wave vectors by Kellerman (1). From the symmetry, these 1000 points are reduced to 48 non-equivalent points at which the vibration frequencies have been obtained by solving the secular determinant. Debye temperature variations at different temperatures by Macfarlane et al. (26) and colossal dielectric constant (CDC) curves for platinum crystals have been computed by using VTBFSM model. By using the sampling technique the corresponding values of Θ_D have been compared with available experimental data (27–29) and the curve for Θ_D vs. absolute temperature (T) was plotted as shown in Figure 1 for platinum. In this temperature range one should take into account the temperature dependence of the frequency spectrum, this requires, however, knowledge of the phonon frequencies at more than one temperature. The variations of Θ_D with specific heats have been used to compute phonon frequencies in the first Brillouin zone and data points for different points were reported in Table II. The effect of anharmonicity is excluded so slight discrepancies between theoretical and experimental results at higher temperatures are seen, though the agreement is almost better with VTBFSM. The calculated (Θ_D–T) curve for platinum has given excellent agreement with the experimental values (8–10). The DOS curve is given in Figure 2 and data points are shown in Table III. The two-phonon Raman spectra are sensitised to the high-frequency side while the specific heats are sensitive to its lower side which is stated the reasonableness of VTBFSM for all wavelength range.

The exploration of model parameters, Debye temperature and DOS are reported by use of the present model VTBFSM for platinum. The conformity with experimental data (8–10) of our theoretical peak is very good for platinum. A successful explanation of spectra has provided the next best test of any model for their higher range of frequency. Small deviations were observed at the higher temperature side due to harmonic approximation in the Debye curve. Better agreement has been obtained with the available experimental data (16–20) and theoretical results. The motivation of this work is the availability of experimental (29) and theoretical (30, 31) work on platinum. Therefore, it may be concluded that the incorporation of VWI is requisite for the absolute interpretation of the phonon dynamical behaviour of platinum. Many researchers have also successfully reported theoretical results for alkali halides (32–42) by use of the present model. Hence, the present model may be understood to provide a powerful and simple approach for a comprehensive study of the harmonic as well as anharmonic elastic properties of the crystals under consideration. The only constraint of VTBFSM is the knowledge of certain experimental parameters needed that can be used as input data.

Platinum and its alloys are very important in industries such as the chemical and glass industries. Some of their main applications are as follows: a functional structural material in the aerospace industry; a catalyst in the nitric acid preparation industry; nozzles for preparing glass fibres; preparing crucibles and utensils that require special properties for use in chemical laboratories. Platinum-rhodium alloys with low rhodium content can also be used as brazing filler metals in metal welding. Although expensive, their high temperature stability and chemical inertia cannot be replaced by other materials (1, 2).

However, the strength and creep resistance of pure platinum would significantly reduce at high temperature because of sharp grain growth. Therefore, it is crucial to find a way to improve the high temperature creep resistance of platinum alloys (3). Many methods of strengthening platinum materials are known. Among these different methods, the best methods are solid solution strengthening and dispersion strengthening. The addition of rhodium has been found to have the best solid solution strengthening effect in high temperature use of platinum. With increasing rhodium content in platinum-rhodium alloys, the normal temperature strength, high temperature endurance strength, creep rupture life and creep activation energy of the alloy increases greatly, so platinum-rhodium alloys are widely used (4, 5). But the high-temperature mechanical properties of platinum-rhodium alloys fail to satisfy the requirement of industrial production with worsening service conditions. Therefore, dispersion-strengthened platinum-rhodium alloys were subsequently developed in Johnson Matthey, UK (6, 7), Engelhard Corporation, USA (8) and Heraeus, Germany (9) from the 1980s to 1990s. They were prepared by adding zirconium, yttrium or scandium oxide particles into the platinum-rhodium alloy under certain process conditions. At present, dispersion-strengthened platinum alloys such as zirconia grain stabilised (ZGS) platinum and oxide dispersion strengthened (ODS) platinum (10, 11), in which Y₂O₃ and ZrO₂ are two commonly used reinforcement particles, are used in industrial applications. According to the experimental results of Hu et al. (12), as the zirconium concentration increases, the platinum-rhodium alloy’s yield stress (0.2% offset) increases significantly, while the elongation decreases slightly. The work hardening rate of particle-reinforced samples increases with the increase of the volume concentration of dispersed particles, which is a typical behaviour of particle dispersion enhancement. In addition, based on a reformulation of the Orowan stress for particle strengthening and by superposing this stress to the matrix stress, a calculated flow stress is in good accord with the experimental value. Though the strength of these particle strengthened platinum alloys can be significantly improved, the further enhancement of the strength is difficult due to the lack of mechanism research and empirical development in the past (12, 13).

Zirconium and yttrium have good plasticity and corrosion resistance. Diffusion strengthening improves the metal strength through adding a second phase or multiple phases and the essence is the interaction of the second phase and the dislocation (13). Nevertheless, the two reinforcement particles, i.e. Y₂O₃ and ZrO₂, are usually thought be to the same as strengthening particles based on Orowan’s equation where the strength increment from the non-deformable dispersed particles is positively correlated with the particle spacing (14, 15). When making samples, we tried to add the yttrium content to 0.1%, and found that the ingot will crack during rolling, so we speculate that different deformation behaviours will have a huge impact on alloy strengthening.

In this study, two kinds of platinum alloys strengthened by Y₂O₃ and ZrO₂, respectively, were prepared and characterised by TEM. A number of differences in deformation behaviour of the two particles during processing were found. These findings are expected to lead to new insights into developing dispersion strengthened high strength platinum alloys.

Nominal composition of the investigated alloys is listed in Table I. The preparation processes have been described in previous research (13) and can be briefly summarised as follows: Pt-20 wt% Rh-0.015 wt% Y without and with 0.1 wt% Zr were smelted in a vacuum induction furnace at 1800–2000°C under argon atmosphere, then cast into a water-cooled copper mould to obtain ingots. The ingots were hot forged to 10 mm plates which were sheared into pieces, followed by mechanical milling to fine powder. The size distribution of the powder was measured using an LS-POP(6) laser particle size analyser (OMEC, China). 97% of the particles were distributed between 5 μm and 60 μm with an average particle size of 25 μm. The powder was sintered and exposed in air to internally oxidise at 1150°C for 4 h. The 30 mm sheets were obtained by hot forging at 1400°C, then rolled to 1 mm sheets at room temperature followed by annealing at 1150°C for 30 min. To prepare the TEM samples, the sheets from rolling-normal direction were mechanically ground to 50–70 μm, followed by argon ion milling at 5 kV until a hole appeared on the centre of the foil. TEM was used to examine the microstructure at 200 kV (JEM-2100 electron microscope (JEOL Ltd, Japan)) and at 300 kV (JEM-3000F field emission electron microscope (JEOL Ltd)). JEM-2100 was used to detect topography of the particles and energy-dispersive X-ray spectroscopy (EDS) was conducted to determine the components; JEM-3000F was used for high-resolution transmission electron microscopy (HR-TEM) observation to determine the structure of the particles. In this paper, Pt-20Rh-0.015Y is referred to as Alloy 1 while Pt-20Rh-0.015Y-0.1Zr is referred to as Alloy 2.

3.1 Thermomechanical Analysis

During preparation processes, especially at high temperature under oxidising atmosphere, the metals in the alloys may be oxidised and the possible oxides are PtO₂, Rh₂O₃, ZrO₂ and Y₂O₃. Their reaction equations are as follows (Equations (i)–(vi):

(i)

(ii)

(iii)

(iv)

(v)

(vi)

According to the Gibbs free energy theorem, a reaction can only occur when the Gibbs free energy is negative and the more negative the Gibbs free energy of a reaction, the more easily the reaction occurs. When the Gibbs free energy is greater than zero, the reaction cannot happen. Figure 1 is the oxidation tendency of platinum, rhodium, zirconium and yttrium, which indicates:

• ΔG of all the reactions increases as temperature increases

• Within the temperature range we calculated, ΔG of Y₂O₃·ZrO₂, Y₂O₃ and ZrO₂ are much smaller than those of Rh₂O₃, Rh₂O and RhO, which means the formation of Y₂O₃·ZrO₂, Y₂O₃ and ZrO₂ are much easier during processing

• ΔG of RhO, Rh₂O and Rh₂O₃ are very close to ΔG = 0 and even greater than 0 at the internal oxidation temperature (1423 K), thus the formation of RhO, Rh₂O and Rh₂O₃ are difficult and can be ignored.

Fig. 1.

The oxidation tendency of platinum, rhodium, zirconium and yttrium

For platinum, when the temperature rises from room temperature, platinum will react with oxygen and form a PtO₂ film on the metal surface. The thickness of PtO₂ would grow with the rise of temperature. When the temperature reaches about 500°C, this process would stop. If the temperature continues to rise, the PtO₂ film will be gradually vaporised. Meanwhile, the higher the temperature, the faster the gasification rate. Samples in this experiment were about 1200°C, the PtO₂ film formed by oxidation had been basically vaporised. Therefore, PtO₂ is not present in the samples.

Overall, according to the thermomechanical analysis, Y₂O₃, ZrO₂ and Y₂O₃·ZrO₂ are easily formed in the platinum-rhodium-(zirconium)-yttrium alloys. Platinum and rhodium are present as almost pure metals.

3.2 Microstructure and Energy-Dispersive X-ray Spectroscopy Analysis of Platinum-Rhodium-(Zirconium)-Yttrium

3.3 Analysis of Deformation Behaviour Between Zirconia and Yttria Particles

In platinum-rhodium-(zirconium)-yttrium alloys, yttrium and zirconium are easily oxidised into Y₂O₃, ZrO₂ and Y₂O₃·ZrO₂ and form corresponding oxides while platinum and rhodium are basically present as pure metals. ZrO₂ and Y₂O₃ particles have been observed in platinum-rhodium-zirconium-yttrium and platinum-rhodium-yttrium, respectively, and verified by EDS analysis and HR-TEM observations. The deformation behaviour of these two oxides is quite different during processing, though they have the same deformation history. The ZrO₂ particles maintained their spherical shape without any visible deformation, while the Y₂O₃ particles were compressed along the normal direction with two cracks forming on two sides of the particles. Insufficient hardness of Y₂O₃ and relatively lower interface strength between Y₂O₃ particles and matrix were supposed to be responsible for deformation of Y₂O₃ during processing.

But it is not for the pgms’ aesthetic or investment value that this collection of papers has been collated, rather to highlight the fascinating science of these incredible metals and their wide range of industrial uses. This special edition of the Johnson Matthey Technology Review will examine both the fundamental properties of these metals and their use in a variety of applications and fields.

The pgms are rare elements, occurring in economic quantities in only a few geographical locations. Demand is generally price inelastic, meaning that consumption volumes are often relatively insensitive to underlying metal prices (1). Importantly, the pgms are widely recovered and recycled; for example, through the recovery of catalytic convertors from end of life vehicles or through a closed-loop system where the catalyst that is installed in a chemical plant is recovered, sent for refining and ultimately reused. Sustainability not only of the metals themselves but also with regard to their end uses is why the pgms are so important, as described in two of the articles – European projects BIORECOVER and PLATIRUS.

The use of platinum, palladium and rhodium is dominated by the automotive sector where for several decades they have been a vital component in emission control catalysts. These three metals have been fundamental in removing carbon monoxide, hydrocarbons and nitrogen oxides from gasoline and diesel engine exhausts to dramatically improve air quality across the world, as detailed in several publications by one of our authors, Martyn Twigg, who in his long career was at the forefront of autocatalyst development (2–4).

In this special edition, Twigg and Emeritus Reader John Burgess from the University of Leicester, UK, have written a two-part commemoration of the late Professor Bob Gillard, discussing his remarkable life, work and contribution to the understanding of transition metal chemistry, particularly the chemistry of rhodium and other platinum group metal complexes.

The use of pgms by the chemical industry is of vital importance to a huge range of bulk and speciality products. One example of this is the oxidation of ammonia to produce nitric acid which has used platinum and rhodium in catalyst gauze for over 100 years (5). The latest work in this field will be discussed by Ashcroft in this special edition. Interestingly, the use of pgm for chemical catalysts has remained one of the more robust areas of demand during the coronavirus disease (COVID-19) pandemic. Nitric acid is used to manufacture both fertilisers for global crop production and explosives for the mining industry, which are essential for the supply of metals such as nickel and platinum and will be central to the future electrification of the automotive fleet.

The minor pgms ruthenium, iridium and osmium can often appear somewhat neglected despite their use in a huge number of applications. Iridium is prized for its high melting point which makes it ideal for use in crucibles to produce high purity metal oxide single crystals, used in medical scanners, light-emitting diode production and surface acoustic wave filters, amongst others. The behaviour and properties of iridium are the subject of two papers in this edition. Osmium is perhaps the least known of the metals given its more limited applications. However, Arblaster has remedied that with a paper discussing the thermodynamic properties of the densest element in the Periodic Table.

The pgms are among the most invaluable elements discovered. To sum up all their useful properties in one key attribute is that they enable the world to be a more sustainable place. Globally we are starting to undergo a monumental change in energy use and production, away from reliance on fossil fuels towards a cleaner, greener, more sustainable model (6). The move to hydrogen as an energy source is vital in the move to a net zero economy, a key example of which is the fuel cell vehicle. Johnson Matthey actually provided the platinum to William Grove when he demonstrated the first fuel cell in 1839 (7). Aside from the use of pgm in the catalyst of the vehicle itself, the production of hydrogen of suitable purity makes use of one of the key properties of palladium. Palladium has an intrinsic selectivity for hydrogen, which makes it an ideal choice for purification membrane technology. In this issue, Faizal et al. discuss the use of palladium in this vital application for the growing hydrogen economy.

The pgms: six of the rarest elements in the Periodic Table that have and continue to change the world around us. Metals that are driving forward sustainable technology and the move towards net zero, metals that will help drive the clean energy revolution, provide food to billions and facilitate communication and data sharing and storage across the globe to enable a more connected society.