Part diagrams information synthesis of extremely ordered intermetallic electrocatalysts: separating alloying and ordering levels

In situ high-temperature X-ray diffraction (HT-XRD) research

We first employed in-situ HT-XRD to grasp the section evolution of PtFe, PtCo, and PtNi in the course of the annealing of their corresponding precursors. Carbon black (Black Pearls 2000, abbreviated as BP2000) supported Pt-M precursor powders had been obtained by the traditional wet-impregnation technique, which had been pressed into sheets for the in-situ HT-XRD measurements underneath the circulation of 5 vol% H₂/N₂. The HT-XRD knowledge had been collected at every temperature in the course of the heating (10 °C min⁻¹), high-temperature holding (850 °C for two h), and cooling levels (10 °C min⁻¹). The small print of the annealing program are proven in Supplementary Fig. 1a, and the outcomes of HT-XRD are summarized in Fig. 2a–c.

a–c In situ HT-XRD outcomes of the Pt–Fe, Pt–Co, and Pt–Ni catalysts, respectively. The depth of the attribute superlattice peak of (110) displays the evolution of ordered fct IMCs. The usual peaks for ordered fct PtFe (JCPDS no. 26-1139), PtCo (JCPDS no. 65-8969), and PtNi (JCPDS no. 65-2797) are additionally proven. Because the experiment proceeded, the precursor sheet shrank to a smaller dimension, ensuing within the emergence of attribute peaks of Al₂O₃ substrates. d–f Calculated ordering levels of PtFe, PtCo, and PtNi in the course of the in-situ HT-XRD experiment, respectively.

For PtFe, we noticed a broad diffraction peak at round 41° when the temperature rose to 300 °C, which corresponded to the (111) airplane of the PtFe alloy. This peak grew to become sharp and shifted to a low 2θ diploma regularly because the annealing temperature elevated, indicating the expansion of the crystal dimension and alloying course of. We additional be aware {that a} weak diffraction peak at round 33° emerged at 800 °C within the heating stage, which corresponded to the (110) superlattice peak, implying that the formation of fct PtFe occurred. The ordering diploma was estimated by evaluating the normalized depth of the (110) peak to the sum of the intensities of the (111) and (200) peaks between the experimental XRD patterns and powder diffraction file (PDF) playing cards. Throughout the high-temperature holding and cooling levels, extra superlattice peaks of the fct PtFe emerged, and their relative intensities elevated, akin to the regularly elevated ordering diploma (Fig. 1d). Within the case of PtCo (Fig. 2b, e), no superlattice peaks had been noticed in the course of the heating and high-temperature holding levels; they began appearing till the temperature went right down to 800 °C in the course of the cooling stage. For PtNi, we didn’t observe any superlattice peaks throughout the entire heating, high-temperature holding, and cooling course of (Fig. 2c, f). In different phrases, we couldn’t receive ordered fct PtNi constructions by standard annealing synthesis.

From the abovementioned in situ HT-XRD research, we summarized the completely different evolution processes of PtFe, PtCo, and PtNi throughout the identical annealing process: (i) for PtFe, the alloying and ordering occurred concurrently within the high-temperature heating stage; (ii) for PtCo, the alloying and ordering occurred individually within the heating/holding levels and cooling stage, respectively; and (iii) for PtNi, solely alloying occurred throughout the entire course of.

The completely different evolution processes of PtFe, PtCo, and PtNi may very well be defined properly by their completely different T_PT proven within the section diagrams (Fig. 1b–d). The T_PT of the PtFe (~1300 °C) was a lot larger than the annealing temperature of the HT-XRD experiments (850 °C). Because of this, the disorder-to-order transition was thermodynamically favorable throughout the entire annealing course of. As soon as the Fe/Pt ratio in some particles approached the stoichiometric values of fct PtFe (i.e., 1:1) in the course of the heating stage, the disordered alloy particles developed into ordered intermetallic constructions concurrently through a thermodynamically pushed section transition. Against this, as T_PT of the PtCo (~830 °C) was a lot decrease than that of the PtFe, the intermetallic PtCo construction was unfavorable on the annealing temperature of 850 °C. On this case, the disorder-to-order transition may very well be realized solely when the pattern was cooled to a decrease temperature under T_PT to enlarge the thermodynamic driving drive for the section transition. The absence of a disorder-to-order transition for the PtNi within the HT-XRD experiments was associated to its considerably low T_PT of ~630 °C, as a result of the low annealing temperature under T_PT was too low to beat the kinetic vitality obstacles of atom ordering.

Synthesis of extremely ordered intermetallic compound (IMC) catalysts

On the idea of the above understanding of the T_PT-dependent structural evolution of PtFe, PtCo, and PtNi, we accordingly optimized the synthesis parameters for every case to realize extremely ordered IMC catalysts with a nominal whole steel content material of ~15 wt% with the BP2000 carbon black helps. The crystal sizes and ordering levels of the catalysts in all syntheses are summarized in Supplementary Desk 1. We first ready the PtFe catalyst (denoted as PtFe-T900-2h) by the identical impregnation with the same annealing program as that used for the HT-XRD experiment, involving the heating (5 °C min⁻¹), high-temperature holding (900 °C for two h), and cooling levels (~9 °C min⁻¹) (Fig. 3a). The ordering diploma of PtFe-T900-2h (22%) was, nonetheless, a lot decrease than that of the pattern obtained within the HT-XRD (68%) (Fig. 2nd). We surmised that the attainable cause was the comparatively longer heating stage within the HT-XRD, which might facilitate the alloying, or the longer cooling stage, and could be useful for the atom ordering. We then up to date the synthesis protocol of PtFe by moreover holding the pattern at a low temperature of 600 °C for two h to spice up the atom ordering. The ready PtFe-T900-2h-T600-6h confirmed an virtually unchanged crystal dimension however elevated ordering diploma (46%). We additionally optimized the synthesis by rising the annealing temperature to 1000 °C to arrange the PtFe-T1000-2h or prolonging the high-temperature holding temperature to six h for getting ready PtFe-T900-6h to advertise the alloying. A excessive ordering of 56% was realized for the PtFe-T1000-2h, however the crystal dimension elevated considerably to 4.1 nm. Due to this fact, PtFe-T900-6h represented the optimum pattern that exhibited the best ordering diploma of 61% and a average crystal dimension of three.1 nm.

a–c X-ray diffraction (XRD) patterns of the Pt–Fe, Pt–Co, and Pt–Ni catalysts, respectively. d Abstract of the annealing protocols, ordering levels, and common particle sizes of PtFe-T900-6h, PtCo-T1000-2h-SC, and PtNi-T1100-2h-T550-12h. e–g Excessive-angular annular dark-field scanning transmission electron microscopy (HAADF-STEM) pictures of PtFe-T900-6h, PtCo-T1000-2h-SC, and PtNi-T1100-2h-T550-12h, respectively. The inserted histograms are the statistics of the particle dimension distributions of the corresponding samples.

For the synthesis of PtCo with a comparatively low T_PT, we first tried low-temperature one-step annealing at 600 °C with a protracted holding time of 6 h to arrange PtCo-T600-6h, however this method failed to realize intermetallic constructions (Supplementary Fig. 2a). This end result revealed that annealing at excessive temperatures above T_PT was needed for alloying Pt with a ample quantity of Co to type disordered constructions with a goal stoichiometric ratio. The pattern (denoted as PtCo-T900-2h) that was ready by one-step annealing at a better temperature of 900 °C for two h confirmed a low ordering diploma of 15% (Fig. 3b). Both prolonging the holding time to six h (denoted as PtCo-T900-6h, with an ordering diploma of 28%) or rising the annealing temperature to 1000 °C (denoted as PtCo-T1000-2h, with an ordering diploma of 37%) may promote the ordering diploma. As a result of all three samples skilled precisely the identical cooling step under T_PT of PtCo for atom ordering, we ascribed the promoted ordering levels of PtCo-T900-6h and PtCo-T1000-2h relative to that of PtCo-T900-2h to the improved alloying diploma of the previous two samples. To additional promote the ordering diploma, we lastly adopted a separate alloying/ordering protocol that concerned a high-temperature annealing step at 1000 °C for two h for alloying, adopted by a really gradual cooling step to 600 °C (−1.1 °C min⁻¹) for ordering to arrange a extremely ordered fct PtCo catalyst (denoted as PtCo-T1000-2h-SC, with an ordering diploma of 63%).

Up to now, there have been only some studies on the synthesis of fct PtNi catalysts, which is probably going due to the a lot decrease T_PT of PtNi than these of PtFe and PtCo. The abovementioned HT-XRD research demonstrated that the intermetallic fct PtNi construction couldn’t be generated in the course of the standard annealing course of. We additionally did not receive fct PtNi catalysts by the traditional one-step impregnation/annealing synthesis at 900, 1000, and 1100 °C for two h (Fig. 3c). It’s value noting that the height place of (111) shifted to a excessive 2θ diploma with the rise within the annealing temperature. By comparability with the usual PDF card (JCPDS:65-2797), it was concluded that the ratio of Ni/Pt in PtNi-T1100-2h approached the goal stoichiometric ratio of fct PtNi. We subsequent moreover adopted a low-temperature-annealing holding stage at 600 °C for six h because the ordering step and adopted a high-temperature-annealing holding stage at 1100 °C for alloying (Supplementary Fig. 2b). Fortuitously, we noticed weak superlattice peaks akin to fct PtNi for the ready PtNi-T1100-2h-T600-6h, regardless of a low ordering diploma of solely 16%. The low ordering diploma was probably attributable to the temperature of the ordering step (~600 °C) being too near T_PT of PtNi (~630 °C), which might result in a weak thermodynamic driving drive for the disorder-to-order section transition, even for the PtNi alloy with the goal stoichiometric ratio of 1:1. Accordingly, we then barely decreased the temperature of the ordering step from 600 to 550 °C, and the ordering diploma of the ready PtNi-T1100-2h-T550-6h catalyst was elevated to twenty-eight% (Supplementary Fig. 2b). We lastly extended the annealing time of the ordering step at 550 °C to 12 h to acquire the optimum pattern of PtNi-T1100-2h-T550-12h with a tremendously promoted ordering diploma of 41%.

Briefly, in line with the abovementioned systematic optimization, we realized the synthesis of extremely ordered PtFe-T900-6h, PtCo-T1000-2h-SC, and PtNi-T1100-2h-T550-12h catalysts by the industrially related impregnation technique (Fig. 3d) (hereafter, denoted as PtFe, PtCo, and PtNi for brief, respectively). We used high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) to investigate the particle dimension distribution of those optimum catalysts, exhibiting common particle sizes of three.5, 4.6, and 5.0 nm for PtFe, PtCo, and PtNi, respectively (Fig. 3e–g), which had been extremely in step with the values calculated by the Debye–Scherer equation primarily based on the full-width at half-maximum of XRD patterns (Supplementary Desk 1). Moreover, aberration-corrected atomic-number(Z)-contrast HAADF-STEM was carried out to investigate the crystal constructions of the intermetallic catalysts on the atomic scale (Fig. 4a–c). For these three catalysts, the HAADF-STEM pictures alongside the [(bar{{{{{{bf{1}}}}}}}{{{{{bf{10}}}}}})] route confirmed the alternating association of Pt and Fe/Co/Ni atom columns, represented by brighter dots and darker dots, respectively (Pt columns had a better depth than Fe/Co/Ni columns with decrease Z values), indicating the presence of the L1₀-type fct intermetallic construction. Quick Fourier rework (FFT) patterns of the corresponding atomic-resolution HAADF-STEM pictures additional verified the presence of fct intermetallic constructions. Vitality-dispersive X-ray spectroscopy (EDS) elemental mapping confirmed the homogeneous distributions of Pt and Fe/Co/Ni within the particular person particles.

Atomic-resolution HAADF-STEM pictures, quick Fourier rework (FFT) patterns, and vitality dispersive X-ray spectroscopy (EDS) elemental mappings of the PtFe (a), PtCo (b), and PtNi (c) IMC catalysts.

Electrochemical efficiency

Previous to the electrochemical assessments, the as-prepared optimum catalysts underwent acid leaching and low-temperature H₂ annealing (400 °C, 2 h) remedies successively to type an electrochemically steady and energetic core/shell constructions consisting of an intermetallic Pt–M core and two to 3 atomic layers of Pt shells^6,17,30,31. Inevitably, the ordering levels of those handled catalysts lower in comparison with the pristine ones due to the lack of transition steel atoms from the floor of the catalysts (Supplementary Fig. 3). We additional famous that the ordering diploma decline for PtFe was extra extreme than that for PtCo and PtNi upon the remedies, which was related to the scale impact. ICP-AES measurements confirmed that the lack of Fe from the PtFe catalyst with smaller common particle dimension was a lot larger than that of PtCo and PtNi catalysts with bigger particle sizes (Supplementary Desk 2).

Rotating disk electrode (RDE) assessments had been first carried out in an O₂-saturated 0.1 M HClO₄ answer at room temperature to judge the ORR efficiency of the handled PtFe, PtCo, PtNi, and business Pt/C catalysts. The PtFe catalyst exhibited the best ORR exercise among the many three intermetallic catalysts, with a big half-wave potential (E_1/2) of 0.936 V, whereas the values had been 0.926, 0.900, and 0.883 V for PtCo, PtNi, and Pt/C (Fig. 5a), respectively. The electrochemical floor areas (ECSA) of the catalysts had been measured by the CO-stripping voltammetry (Supplementary Fig. 5). The PtFe catalyst confirmed a excessive ECSA of 85.4 m² g^–1, which was barely decrease than that of Pt/C (111.6 m² g^–1) (Supplementary Desk 2). When it comes to the mass exercise (MA) and particular exercise (SA), PtFe exhibited a excessive MA and SA of two.61 A mg⁻¹ and three.01 mA cm⁻² at 0.9 V, which had been corresponding to these of PtCo (MA, 2.10 A mg_Pt⁻¹; SA, 3.87 mA cm⁻²) and better than these of PtNi (MA, 0.53 A mg_Pt⁻¹; SA, 2.05 mA cm⁻²) and Pt/C (MA, 0.34 A mg_Pt⁻¹; SA, 0.30 mA cm⁻²) (Fig. 5b). Electrochemical impedance spectroscopy (EIS) analyses confirmed a decrease cost switch resistance in PtFe catalyst, confirming its superior exercise (Supplementary Fig. 6). We attributed the best MA of PtFe to its larger ECSA and ordering diploma in contrast with these of PtCo and PtNi. The sturdiness of the PtFe and Pt/C catalysts was evaluated by conducting an accelerated sturdiness take a look at (ADT) through the RDE. The PtFe catalyst exhibited a slight downshift of 9 mV for E_1/2 after 30,000 ADT cycles (Fig. 5c). The MA and SA of the PtFe maintained 70% and 78% of the preliminary values after 30,000 ADT cycles, respectively (Fig. 5d). Against this, the Pt/C catalyst confirmed a 50% drop of the MA after the ADT (Supplementary Fig. 8 and Supplementary Desk 4).

a Oxygen discount response (ORR) polarization curves of the Pt/C, PtFe, PtCo, and PtNi catalysts (with inner resistance corrected). b Comparability of mass actions (MAs) and particular actions (SAs) of the catalysts at 0.9 V (versus reversible hydrogen electrode (RHE)). The SAs had been normalized by the electrochemical floor areas (ECSAs) estimated from CO stripping. c ORR polarization curves of the PtFe catalyst earlier than and after accelerated sturdiness take a look at (ADT). d MAs and SAs of PtFe catalysts earlier than and after ADT. e H₂–air single-cell polarization curves of PtFe and Pt/C cathodes. f MA in H₂–O₂ take a look at and voltage at 0.8 A cm^–2 in H₂–air take a look at of PtFe cathode earlier than and after ADT.

We additional carried out membrane electrode meeting (MEA) assessments to judge the catalyst performances in sensible PEMFCs. The MAs of the catalysts had been first evaluated at 0.9 V_iR-correct in H₂–O₂ cell assessments at 80 °C. The MA of the PtFe reached 0.96 A mg⁻¹, which was larger than that of Pt/C (0.20 A mg⁻¹). For H₂–air single-cell assessments, the PtFe cathode exhibited a present density of 318 mA cm⁻² at 0.8 V within the kinetic area (Fig. 5e), which was larger than that of the Pt/C cathode (205 mA cm⁻²) with the identical loading. As well as, The PtFe cathode with a low-Pt loading of 0.055 mg_Pt cm⁻² exhibited a comparable energy density to that of the Pt/C cathode with a excessive Pt loading of 0.20 mg_Pt cm⁻² within the excessive present density area, at which the MEA with low-Pt loading suffered from a better local-O₂ switch resistance³². We additional evaluated the sturdiness of PtFe by making use of 30,000 cycles of sq. waves from 0.6 to 0.95 V in MEA. Notably, the PtFe catalyst retained 84% of its MA in H₂–O₂ assessments and confirmed a voltage lack of lower than 15 mV at 0.8 A cm⁻² in H₂–air assessments after ADT (Fig. 5f and Supplementary Fig. 9). ICP-AES measurement was carried out to judge the Fe loss in the course of the ADT. The Fe/Pt ratio declined barely from 0.41 o 0.36 after the ADT, indicated the excessive structural stability of the intermetallic catalysts⁵.