Processing of ZrB2 tribo-ceramics by reactive spark plasma sintering of ZrH2+2B subjected to high-energy pre-ball-milling

Journal of the European Ceramic Society 43 (2023) 5195–5206 Available online 23 April 2023 0955-2219/© 2023 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/). Processing of ZrB 2 tribo-ceramics by reactive spark plasma sintering of ZrH 2 +2B subjected to high-energy pre-ball-milling Jesús L´ opez-Arenal a , Bibi Malmal Moshtaghioun a , * , Diego G´ omez-García a , Angel L. Ortiz b , * a Departamento de Física de la Materia Condensada, Universidad de Sevilla, 41080 Sevilla, Spain b Departamento de Ingeniería Mec´ anica, Energ´ etica y de los Materiales, Universidad de Extremadura, 06006 Badajoz, Spain ARTICLE INFO Keywords: ZrB 2 Reactive spark plasma sintering High-energy ball-milling Tribo-ceramics Sliding wear ABSTRACT The reactive spark plasma sintering (RSPS) of monolithic ZrB 2 ceramics from ZrH 2 +2B powder mixtures subjected to shaker pre-milling was investigated, and compared with other three sintering approaches. It was found that RSPS is optimal at 1850 ◦C, which results in fully-dense ZrB 2 ceramics with ~20 GPa hardness. Comparatively, at 1850 ◦C RSPS from the simply-mixed ZrH 2 +2B powder mixture, SPS from the commercial ZrB 2 powder, and SPS from the shaker-milled ZrB 2 powder result in non-dense (76.7–86.7%) and softer (6.0–11.8 GPa) ZrB 2 ceramics. Furthermore, the optimally RSPS-ed ZrB 2 ceramic was subjected to unlubricated slidingwear tests against diamond under 40 N load for 1000 m of sliding, demonstrating that it is a promising triboceramic that only undergoes mild tribo-oxidative wear at 10 –8 mm 3 /(N⋅m) in the form of a slight plasticitydominated two-body abrasion with eventual formation and partial loss of a self-lubricating and protective oxide tribolayer. 1. Introduction ZrB 2 is a member of the ultra-high-temperature ceramic (UHTC) family [1–5], so called in virtue of their high melting points above 3000 ◦C, and doubtless the most investigated of them all. Indeed, over the last two decades a substantial knowledge base has been accumulated on the synthesis, sintering, thermal and mechanical properties, and oxidation of monolithic ZrB 2 ceramics and ZrB 2 –based composites [6,7], especially with a view to their possible use in the field of extreme-environment engineering and, more particularly, in aerospace engineering (e.g., hypersonic flight, atmospheric re-entry, and scramjet and rocket propulsion) [8]. However, they could also be used as substitutes for other structural ceramics in many other applications, particularly in the field of tribology. Although not as hard as other structural ceramics (e.g., B 4 C and SiC), these UHTCs are also very hard [9], which in principle makes them suitable for applications requiring wear resistance (e.g., bearings, valves, and nozzles) [10]. Their ultra-refractoriness and high thermal conductivity are in fact a plus for resisting and alleviating, respectively, frictional heating. Also, their high electrical conductivity allows them to be used as substitutes for metals in tribological applications requiring electrical contact, and to be machined into complex shapes by electrical discharge machining rather than with diamond tools. Not surprisingly, there have already been studies on the wear behaviour of monolithic ZrB 2 ceramics and ZrB 2 –based composites under varied tribological conditions (e.g., loads, tribo-part, and medium) and contact geometries (e.g., sliding and scratch) [11–24]. However, a limiting factor for the widespread use of these UHTCs in tribological applications is their poor densification capability because porosity is a major microstructural source of degradation of the wear resistance and other engineering properties of ceramics. Indeed, they are extremely difficult to sinter, more as monolithic than as composites, so they are more often fabricated by hot pressing (HP) and spark plasma sintering (SPS) than by pressureless sintering (PS). Importantly, reactive sintering (RS), in which a mixture of elemental powders reacts and densifies during the sintering cycle itself, has also been used as an alternative to the classical sintering (PS, HP, or SPS) of ZrB 2 powders in the case of monoliths, or of powder mixtures of ZrB 2 with other ceramics in the case of composites. Table 1 lists previous work on the RS of monolithic ZrB 2 ceramics [25–28]. Despite it being difficult to draw conclusions from the comparison of these few studies, it seems that RS of ZrB 2 benefits from the application of pressure and fast heating (i.e., the utilisation of SPS), the use of ZrH 2 reactant instead of Zr reactant, and the shaker pre-milling treatment. Also importantly, RS has been shown * Corresponding authors. E-mail addresses: [email protected] (B.M. Moshtaghioun), [email protected] (A.L. Ortiz). Contents lists available at ScienceDirect Journal of the European Ceramic Society journal homepage: www.elsevier.com/locate/jeurceramsoc https://doi.org/10.1016/j.jeurceramsoc.2023.04.047 Received 18 March 2023; Received in revised form 10 April 2023; Accepted 20 April 2023 Journal of the European Ceramic Society 43 (2023) 5195–5206 5196 to be comparatively more effective in producing dense monolithic UHTCs than more typical sintering approaches [26,29]. Given the above context, the present study was undertaken with three novel objectives in mind. The first objective was to investigate the RSPS of ZrB 2 from a ZrH 2 +2B powder mixture subjected to shaker premilling, which has not yet been explored despite the few earlier studies [25–28] suggesting that this would be the ideal combination of reactants, pre-treatment, and sintering technique. The second objective was to compare under a common experimental platform the RSPS of ZrB 2 from the shaker-milled ZrH 2 +2B powder mixture with RSPS of ZrB 2 from the simply-mixed ZrH 2 +2B powder mixture, SPS of a typical commercial ZrB 2 powder, and SPS of the commercial ZrB 2 powder subjected to shaker pre-milling. Not only has this comparison between the four not yet been done, but also the two existing partial comparisons either took data from different studies [26] or were not for ZrB 2 (but for HfB 2 and TaB 2 ) [29]. The third objective was to investigate the unlubricated sliding-wear behaviour of the optimally RSPS-ed ceramic against diamond (to thus simulate the scenario of wear against a much harder tribo-part), which involves fundamental tests not as yet performed but necessary to assess the potential usefulness of these ceramics in tribological applications. Table 1 Summary of conditions and results of previous studies on the RS of monolithic ZrB 2 ceramics. Reference Sintering technique Powder mixture composition Powder mixture preparation Sintering conditions Densification degree [25] pressureless reactive sintering Zr+2B shaker milling 1 h/2000 ◦C 72% [25] pressureless reactive sintering Zr+2B shaker milling 1 h/2200 ◦C 79% [25] pressureless reactive sintering Zr+2B simply mixed 1 h/2200 ◦C 70% [26] reactive hot pressing ZrH 2 +2B simply mixed 2 h/1900 ◦C/50 MPa 98.9% [26] reactive hot pressing ZrH 2 +2B simply mixed 1 h/2050 ◦C/50 MPa 99.6% [26] reactive hot pressing ZrH 2 +2B simply mixed 15 min/2100 ◦C/50 MPa 99.2% [27] reactive spark plasma sintering Zr+2.1B shaker milling 20 min/1900 ◦C/20 MPa/500 ◦Cmin –1 87% [27] reactive spark plasma sintering Zr+2.1B shaker milling 20 min/1900 ◦C/50 MPa/500 ◦Cmin –1 95% [27] reactive spark plasma sintering Zr+2.1B shaker milling 20 min/2000 ◦C/20 MPa/200 ◦Cmin –1 96% [28] reactive spark plasma sintering ZrH 2 +2B wet milling 10 min/1800 ◦C/50 MPa/100 ◦Cmin –1 97.2% Table 2 Set of ceramics fabricated and their fabrication conditions. Sintering technique Powder mixture used Sintering temperature a Number reactive spark plasma sintering shaker-milled ZrH 2 +2B 1500–1900 ◦C in 50 ◦C steps 9 reactive spark plasma sintering simply-mixed ZrH 2 +2B 1850 ◦C 1 spark plasma sintering commercial ZrB 2 1850 ◦C 1 spark plasma sintering shaker-milled commercial ZrB 2 1850 ◦C 1 a Other sintering conditions are 100 ◦C/min heating ramp, 5 min soaking at peak temperature, and 50 MPa pressure. Fig. 1. FE-SEM images of the starting (A) ZrH 2 and (B) B powders as well as of (C-D) the ZrH 2 +2B powder mixture subjected to shaker milling. Imaging of the powder mixture was done with backscattered electrons to differentiate between ZrH 2 (light particles) and B (dark particles). J. L´ opez-Arenal et al. Journal of the European Ceramic Society 43 (2023) 5195–5206 5197 2. Experimental procedure 2.1. Processing and characterisations Commercially available (MaTecK, Germany) powders of ZrH 2 (99.7% purity) and B (95% purity) were purchased, combined in molar ratio 1ZrH 2 :2B, and subjected to dry high-energy ball-milling for 1 h in a shaker mill (Spex D8000, Spex CertiPrep, USA) at 1060 back-and-forth cycles/min using a hardened-steel container with WC balls (6.7 mm in diameter) under a ball-to-powder weight ratio of 4, in an inert atmosphere of ultrahigh-purity Ar to prevent oxidation of the powder mixture during milling. Also, a first sacrificial milling was performed, which has been shown to be effective in preventing the contamination of the powder mixture by the milling tools [30–35]. The shaker-milled ZrH 2 +2B powder mixture was subsequently characterised using field-emission scanning electron microscopy (FE-SEM; FEI Teneo, FEI Inc., USA), transmission electron microscopy (TEM; Tecnai G 2 20 Twin, FEI Inc., The Netherlands), X-ray diffractometry (XRD; D8 Advance, Bruker AXS, Germany) under CuK α incident radiation, and thermogravimetry (TG; Labsys Evo, Setaram, France) coupled with differential thermal analysis (DTA; Labsys Evo, Setaram, France) under flowing Ar, and compared with the as-purchased ZrH 2 and B powders and with the simply-mixed ZrH 2 +2B powder mixture. The shaker-milled ZrH 2 +2B powder mixture was then loaded into graphite dies (2.5 cm diameter) lined with graphite foils and covered by graphite blankets, and densified using an SPS furnace (HP-D-10, FCT Systeme GmbH, Germany) operated in dynamic vacuum at target temperatures in the range 1500–1900 ◦C (as measured by an axial pyrometer and reached at 100 ◦C/min), without intermediate holds, for 5 min under 75 MPa pressure (applied at 300 ◦C). The resulting ceramics were ground and diamond-polished to a 0.25- μ m finish using conventional ceramographic procedures, and were characterised microstructurally by XRD, SEM (S-3600 N, Hitachi, Japan) on fracture surfaces and, selectively, FE-SEM on polished and electrochemically-etched surfaces (45 s at 0.4 mA/cm 2 in 0.05 M KOH solution), and water immersion porosimetry (i.e., the Archimedes method). Also, their hardness was measured by Vickers indentation (Duramin, Struers A/S, Denmark). 10 indentations were performed for each material, at 9.8 N load, 40 µm/s load rate, and 20 s dwell time. As reference for comparison, three additional powder batches were also prepared and RSPS-ed/SPS-ed. Two were a typical commercially available ZrB 2 powder (Grade B, H. C. Starck, Germany) used both in its as-received condition and shaker pre-milled under identical conditions as the ZrH 2 +2B powder mixture. The third was the simply-mixed ZrH 2 +2B powder mixture, which was homogenised in ethanol for 24 h and conveniently dried and deagglomerated. These three powder batches were RSPS-ed/SPS-ed under the optimal RSPS temperature of 1850 ◦C previously identified for the shaker-milled ZrH 2 +2B powder mixture. Table 2 lists the 4 types of monolithic ZrB 2 ceramics fabricated Fig. 2. TEM images of the ZrH 2 +2B powder mixture subjected to shaker milling, taken within the agglomerates at (A-B) different locations and magnifications. Fig. 3. XRD patterns of the simply-mixed and shaker-milled ZrH 2 +2B powder mixtures. Peak assignations, as identified using the PDF2 database (PDF number 00–036–1339 for ε -ZrH 2 ), are included. The intensity scale is the same for the two XRD patterns to facilitate their direct visual comparison. Fig. 4. Curves of TG and DTA (exothermic upwards) of the shaker-milled ZrH 2 +2B powder mixture. The two curves were registered simultaneously as a function of the temperature in the range 25–1600 ◦C under heating at 10 ◦C/ min in flowing Ar atmosphere. J. L´ opez-Arenal et al. Journal of the European Ceramic Society 43 (2023) 5195–5206 5198 Fig. 5. XRD patterns of the nine ceramics fabricated by RSPS (for 5 min under 75 MPa) from the shaker-milled ZrH 2 +2B powder mixture. The RSPS temperature is indicated. Peak assignations, as identified using the PDF2 database (PDF numbers 00–034–0423 for ZrB 2 , 00–037–1484 for m-ZrO 2 , and 00–050–1089 for t-ZrO 2 ), are included. The letters m and t mean monoclinic and tetragonal, respectively. The intensity scale is logarithmic to facilitate observation of the weaker peaks. Very weak ZrO 2 peaks have not been labelled. J. L´ opez-Arenal et al. Journal of the European Ceramic Society 43 (2023) 5195–5206 5199 Fig. 6. SEM images of the fracture surface of the nine ZrB 2 ceramics fabricated by RSPS (for 5 min under 75 MPa) from the shaker-milled ZrH 2 +2B powder mixture at (A) 1500 ◦C, (B) 1550 ◦C, (C) 1600 ◦C, (D) 1650 ◦C, (E) 1700 ◦C, (F) 1750 ◦C, (G) 1800 ◦C, (H) 1850 ◦C, and (I) 1900 ◦C. The magnification is the same to facilitate the direct visual comparison of microstructural features (porosities and grain sizes). J. L´ opez-Arenal et al. Journal of the European Ceramic Society 43 (2023) 5195–5206 5200 and their fabrication conditions. 2.2. Unlubricated sliding-wear tests and characterisations The optimal of all ceramics fabricated was subjected to unlubricated sliding wear, under ambient conditions. Specifically, sliding-wear tests were done (MFT-3000, Rtec Instruments, USA), in duplicate, in the ballon-disk configuration without external lubricant at 40 N load, 10 cm/s linear sliding speed, 2 mm track radius, and 1000 m total sliding distance, using 6.02 mm diameter diamond-coated SiC balls (Dball G10, Nova Diamant, UK) as counter-parts. The coefficient of friction (CoF) was continuously logged during the wear tests. The worn surfaces of the ZrB 2 ceramic were examined using different techniques. Specifically, digital optical microscopy (DOM; AM7915MZT-Edge, Dino-Lite, The Netherlands), optical microscopy (OM; Epiphot 300, Nikon, Japan), and optical profilometry (OP; Profilm 3D, Filmetric, USA) in white light interferometry mode were used to inspect the wear-induced macrodamage. OP was additionally used to compute the worn volume, and thence the specific wear rate (SWR) and the attendant wear resistance. Moreover, SEM together with energy-dispersive X-ray spectroscopy (EDS; flash Detector 3001, R¨ ontec GmbH, Germany) were used to inspect the wear-induced micro-damage. The combination of DOM, OM, OP, and SEM/EDS allowed the wear mode and mechanism(s) to be identified. 3. Results and discussion 3.1. Characterisation of the shaker-milled powder mixture Fig. 1 shows FE-SEM images of the as-received ZrH 2 and B powders as well as of the shaker-milled ZrH 2 +2B powder mixture. It can be seen in Fig. 1A that the ZrH 2 powder comprises rounded particles with sizes ranging from less than one micron to more than ten microns, and in Fig. 1B that the B powder is finer and more uniform, with an apparent particle size in the submicrometric range. Comparatively, it can be seen in Fig. 1C-D that the shaker-milled ZrH 2 +2B powder mixture is formed by micrometric agglomerates of ultrafine particles that are one and two orders of magnitude smaller than the B and ZrH 2 starting particles, respectively. The FE-SEM images of Fig. 1C-D, taken with backscattered electrons to have compositional contrast, also show a uniform dispersion of the two types of particles within the agglomerates (i.e., ZrH 2 (lighter particles) and B (darker particles)), indicating that shaker milling, which is the most effective form of high-energy ball-milling [36], not only refined the particles’ sizes, but also intimately mixed the ZrH 2 and B powders. Fig. 2 shows TEM images of the shaker-milled ZrH 2 +2B powder mixture revealing that the agglomerates are compact but porous (Fig. 2A), and confirming that they contain ultrafine and even nanometric particles (Fig. 2B), indicative that they formed through short thermal excursion in the presence of pressure. This is because at the Fig. 7. Values of (A) relative density and (B) hardness as a function of their sintering temperature for the twelve ceramics fabricated. The dotted and dashed lines as well as the coloured boxes in (A) distinguish the regimes of initial, intermediate, and final sintering (i.e., densification lower than 70%, between 70% and 90%, and more than 90%, respectively). Errors are smaller than the point size. The solid lines through the data are guides for the eye. Fig. 8. FE-SEM images of the (A) ZrH 2 +2B powder mixture prepared simply by wet homogenisation, (B) a typical commercial ZrB 2 powder, and (C) the commercial ZrB 2 powder subjected to the same shaker milling as the ZrH 2 +2B powder mixture. Imaging of the powder mixture was done with backscattered electrons to differentiate between ZrH 2 (light particles) and B (dark particles). J. L´ opez-Arenal et al. Journal of the European Ceramic Society 43 (2023) 5195–5206 5201 collision site there is a local temperature spike of ~300 ◦C with the presence of high compressive stresses of ~6 GPa of very short duration of the order of 10 –6 –10 –5 s [37]. Fig. 3 shows the XRD patterns of both the simply-mixed and shakermilled ZrH 2 +2B powder mixtures. It is clear that the two have the same phase composition, with only crystalline ε -ZrH 2 plus amorphous B. Therefore, the absence of ZrB 2 in the shaker-milled ZrH 2 +2B powder mixture rules out the occurrence of reaction between ZrH 2 and B during high-energy ball-milling. The two XRD patterns differ however in that the diffraction peaks are much broader and less intense after shaker milling, which is further evidence of particle size refinement and mechanical activation during high-energy ball-milling. Also importantly, the absence of WC peaks in the XRD pattern after shaker milling confirms the non-contamination of the ZrH 2 +2B powder mixture by the milling tools, attributable to the first sacrificial milling depositing a layer of powder mixture on the WC balls and hardened-steel container [38]. Fig. 4 shows the TG–DTA curves of the shaker-milled ZrH 2 +2B powder mixture registered at 10 ◦C/min in the temperature range 25–1600 ◦C in flowing Ar. It can be seen in the TG curve that the total mass loss is only ~2% wt%, and ~1.8% wt% when the initial loss of physisorbed water is excluded, and that this occurs during the first ~850 ◦C. This mass loss is simply due to the ZrH 2 decomposition because the reaction ZrH 2 →Zr+H 2 (g) predicts the release of ~2.16 wt% of hydrogen gas that would imply a mass loss of ~1.75 wt% in the Fig. 9. XRD patterns of the simply-mixed ZrH 2 +2B powder mixture, the commercial ZrB 2 powder, and the shaker-milled ZrB 2 powder, as indicated. Peak assignations, as identified using the PDF2 database (PDF numbers 00–036–1339 for ε -ZrH 2 and 00–034–0423 for ZrB 2 ), are included. The intensity scale is the same for the XRD patterns of the two ZrB 2 powders to facilitate their direct visual comparison. Very weak ZrO 2 peaks have not been labelled. Fig. 10. XRD patterns of the three reference ceramics fabricated by RSPS at 1850 ◦C (for 5 min under 75 MPa) from the simply-mixed ZrH 2 +2B powder mixture and by SPS at 1850 ◦C (for 5 min under 75 MPa) from both the commercial ZrB 2 powder and the shaker-milled ZrB 2 powder, as indicated. Peak assignations, as identified using the PDF2 database (PDF numbers 00–034–0423 for ZrB 2 , 00–037–1484 for m-ZrO 2 , and 00–050–1089 for t-ZrO 2 ), are included. Very weak ZrO 2 peaks have not been labelled. Fig. 11. SEM images of the fracture surface of the three reference ZrB 2 ceramics fabricated (A) by RSPS at 1850 ◦C (for 5 min under 75 MPa) from the simply-mixed ZrH 2 +2B powder mixture and by SPS at 1850 ◦C (for 5 min under 75 MPa) from (B) the commercial ZrB 2 powder and from (C) the shaker-milled ZrB 2 powder. The magnification is the same and equal to that in Fig. 6 to facilitate the direct visual comparison of microstructural features (porosities and grain sizes). J. L´ opez-Arenal et al. Journal of the European Ceramic Society 43 (2023) 5195–5206 5202 ZrH 2 +2B powder mixture. The ZrH 2 decomposition is known to occur in various steps [39], which is also consistent with the shape of the TG curve. On the other hand, it can be seen in the DTA curve that two broad exothermic events occurred, the first in the temperature range ~750–1100 ◦C without a well-defined maximum, and the second from ~1200 ◦C onwards with a maximum at ~1500 ◦C. The former event comes accompanied by mass loss, and is thus due to the ZrH 2 decomposition (ZrH 2 →Zr+H 2 (g)). The latter event does not come accompanied by mass loss and releases much more heat, and is therefore attributable to the ZrB 2 formation reaction (Zr+2B→ZrB 2 ). This, which is consistent with thermodynamic data [40], suggests that ZrB 2 should be the main phase in the RSPS-ed ceramics. 3.2. Optimisation of the RSPS temperature Fig. 5 shows the XRD patterns of the 9 ceramics fabricated by RSPS from the shaker-milled ZrH 2 +2B powder mixture at 1500–1900 ◦C for 5 min under 75 MPa pressure. It can be seen that they all contain essentially only ZrB 2 , indicating that the expected reaction ZrH 2 +2B→ZrB 2 was completed at an RSPS temperature below 1500 ◦C. There are also traces (i.e., <5 vol%) of ZrO 2 , which is attributable to the slight spontaneous passivation of ultrafine ZrH 2 particles during placement of the dies in the chamber of the SPS furnace. In any case, what matters is that the reaction of ZrB 2 formation was completed at a lower temperature during RSPS than expected from the DTA curve in Fig. 4, which is not a surprise for two reasons. Firstly, at the heating rates used in both RSPS and DTA the reaction is governed by solid-state diffusion [27], and the reaction kinetics is therefore favoured by the faster heating and the pressure application as there is less particle growth and more intimate particle contact. Secondly, the actual temperature in the sample during RSPS is in fact higher than that measured by the axial optical pyrometer [41,42]. Fig. 6 shows SEM images of the fracture surface of the entire set of ZrB 2 ceramics fabricated by RSPS from the shaker-milled ZrH 2 +2B powder mixture. The microstructural changes with increasing RSPS temperature are evident both in terms of degree of densification and of grain shape and size. Thus, it can be seen that the two ceramics RSPS-ed at 1500 ◦C and 1550 ◦C simply look like compacted rounded fine powder, that the two ceramics RPS-ed at 1600 ◦C and 1650 ◦C are very porous and have relatively faceted fine grains, that the three ceramics RSPS-ed at 1700 ◦C, 1750 ◦C, and 1800 ◦C are still porous, but with rather isolated pores and already with fine–medium faceted grains, and that the two ceramics RSPS-ed at 1850 ◦C and 1900 ◦C are already essentially dense and with medium and coarse faceted grains, respectively. These SEM observations indicate that 1850 ◦C is the optimal RSPS temperature for the shaker-milled ZrH 2 +2B powder mixture. Note that 1800 ◦C has been reported for the RSPS of a wet-milled ZrH 2 +2B powder mixture [28], but measured with a radial pyrometer which underestimates the actual temperature of conductive samples relative to an axial pyrometer [41,42]. Fig. 7 shows the relative density (Fig. 7A), as measured by the Archimedes method, and the hardness (Fig. 7B), as measured by Vickers indentation, of all ZrB 2 ceramics quantitatively confirming the above conclusion that RSPS is optimal at 1850 ◦C. Certainly, it can be seen in Fig. 7 that both the degree of densification and the hardness increase with increasing RSPS temperature, and that the latter scales directly Fig. 12. FE-SEM images of the polished and electrochemically-etched surface of the ZrB 2 ceramic fabricated by RSPS at 1850 ◦C for 5 min under 75 MPa from the shaker-milled ZrH 2 +2B powder mixture, taken at (A-B) different locations and magnifications. The larger, darker grains are ZrB 2 , and the smaller, lighter grains are ZrO 2 . Fig. 13. Results deriving from the sliding-wear tests for the ZrB 2 ceramic fabricated by RSPS at 1850 ◦C (for 5 min under 75 MPa) from the shaker-milled ZrH 2 +2B powder mixture. (A) Friction curves measured as a function of the distance slid, (B) 3-D image representative of the residual wear track at the conclusion of the wear tests, as obtained by OP, and (C) typical 2-D crosssectional profiles extracted from zones with more and less damage of the wear tracks. The arrow in (B) denotes the sliding direction. J. L´ opez-Arenal et al. Journal of the European Ceramic Society 43 (2023) 5195–5206 5203 with the former. Specifically, it can be seen that RSPS of the shakermilled ZrH 2 +2B powder mixture yields ZrB 2 ceramics having reached the boundary between the initial and intermediate sintering regimes at 1500 ◦C and 1550 ◦C (∼70% dense with <6 GPa hardness), the middle stage of the intermediate sintering regime at 1600 ◦C (∼82% dense with ∼10 GPa hardness), the later stage of the intermediate sintering regime at 1650 ◦C (∼88% dense with ∼13 GPa hardness), the earlier stage of the final sintering regime at 1700 ◦C (∼92% dense with ∼17 GPa hardness), the middle stage of the final sintering regime at 1750 ◦C and 1800 ◦C (∼95% dense with ∼18–19 GPa hardness), and finally the later stage of the final sintering regime at and above 1850 ◦C (∼98.5% dense with ∼20 GPa hardness). Consequently, it is clear that RSPS at 1850 ◦C maximises hardness, and that RSPS temperatures above 1850 ◦C are no longer beneficial in terms of hardness and, however, are detrimental in terms of microstructural coarsening (which in turn is known to degrade the wear resistance of polycrystalline ceramics [10,43–49]). 3.3. Comparison with RSPS from ZrH 2 +2B without pre-milling and with SPS from ZrB 2 To univocally compare the RSPS from the shaker-milled ZrH 2 +2B powder mixture with other sintering approaches, three reference ceramics were also fabricated at 1850 ◦C for 5 min under 75 MPa pressure, one by RSPS from the simply-mixed ZrH 2 +2B powder mixture and two by SPS of a typical commercial ZrB 2 powder both without and with the same shaker pre-milling as the ZrH 2 +2B powder mixture. Figs. 8 and 9 show FE-SEM images and the XRD patterns of these three powders, respectively, showing that the first comprises a mixture of fine+coarse crystalline ZrH 2 particles (∼1–15 µm) and submicrometric (<1µm) amorphous B particles, that the second comprises micrometric ZrB 2 particles (∼2–3µm, as indicated by the manufacturer), and that the third comprises micrometric agglomerates of ultrafine ZrB 2 particles plus micrometric particles not refined during shaker milling. Therefore, despite the large ZrH 2 particle size, the ZrH 2 +2B powder mixture milled better than the ZrB 2 powder, probably because ZrH 2 is more brittle than ZrB 2 . Figs. 10 and 11 show the XRD patterns and SEM images of the fracture surface of the three resulting ceramics, respectively. It can be seen in Fig. 10 that the ceramic RSPS-ed from the simply-mixed ZrH 2 +2B powder mixture only contains ZrB 2 , indicating that 1850 ◦C is a sufficiently high temperature to also complete the reaction of ZrB 2 formation from coarse particulate reactants. The ceramics SPS-ed from the commercial and shaker-milled ZrB 2 powders have the expected phase compositions, the former only with ZrB 2 and the latter with ZrB 2 plus traces of ZrO 2 impurities. The existence of ZrO 2 is again attributable to the slight spontaneous passivation of the ultrafine/fine ZrB 2 particles prior to SPS. Interestingly, the fact that the ZrH 2 +2B powder mixture milled better than the ZrB 2 powder, and therefore that its greater specific surface area must have resulted in greater formation of passivating oxide, together with the existence of only tetragonal ZrO 2 impurities in the ceramic RSPS-ed at 1850 ◦C from the former but both monoclinic and tetragonal ZrO 2 impurities in the ceramic SPS-ed at 1850 ◦C from the latter, suggest that the H 2 released during the ZrH 2 decomposition could have partially reduced the passivating oxide layers. More importantly, it can be seen in Fig. 11 that the three are, to a greater or lesser extent, very porous and with coarse-grained microstructures. Of the three, the one SPS-ed from the commercial ZrB 2 powder is the least dense and coarse-grained, attributable to the very poor densifiability of micrometric ZrB 2 [6,7]. The one SPS-ed from the shaker-milled ZrB 2 powder is the densest, but not the coarsest grained, attributable to the enhanced sinterability of refined ZrB 2 [50,51]. Lastly, the one RSPS-ed from the simply-mixed ZrH 2 +2B powder mixture is the coarsest grained, but not the densest, attributable to the very large ZrH 2 particle size. Their relative densities measured by the Archimedes method, which have also been included in Fig. 7A for the sake of comparison, are as low as ~83.3% for the one RSPS-ed from the simply-mixed ZrH 2 +2B powder mixture, ~76.7% for the one SPS-ed from the commercial ZrB 2 powder, and ~86.7% for the one SPS-ed from shaker-milled ZrB 2 powder. Therefore, the three have reached only the intermediate sintering regime, but the first its middle stage, the second its early stage, and the third its later stage according to the classification used above. Also, it can be seen in Fig. 7B that, owing to their low degrees of densification, these three ZrB 2 ceramics are comparatively much softer than the counterpart RSPS-ed at 1850 ◦C from the shaker-milled ZrH 2 +2B powder mixture (~11.1, 6.0, and 11.8 GPa, respectively, vs 20 GPa). Therefore, it is evident that RSPS from the shaker-milled ZrH 2 +2B powder mixture is the best choice of the four processing approaches. Indeed, comparatively it achieved at 1600 ◦C, ~1575 ◦C, and 1650 ◦C the same degrees of densification as RSPS at 1850 ◦C from the simplymixed ZrH 2 +2B powder mixture and SPS at 1850 ◦C from both the commercial and shaker-milled ZrB 2 powders, respectively. Importantly, neither RSPS nor high-energy ball-milling alone is sufficient to achieve Fig. 14. (A-B) OM images of the residual wear tracks in the ZrB 2 ceramic fabricated by RSPS at 1850 ◦C (for 5 min under 75 MPa) from the shaker-milled ZrH 2 +2B powder mixture. (C) DOM images of the uncleaned and cleaned wear tracks. The arrows denote the sliding direction. J. L´ opez-Arenal et al.