Highly active nano-sized iridium catalysts: synthesis and operando spectroscopy in a proton exchange membrane electrolyzer

Ultra-high purity nano-sized iridium enclosed in a monolayer of IrIII/IrIV oxides/hydroxides leads to an enhancement in OER activity.

: XRD patterns of Ir-nano 99.8, Ir-nano 99.5, Ir-nano 99.5\CTAB, Ir-nano H 2 O catalysts. X-ray diffraction data were collected by using a D8 Discover GADDS diffractometer with VÅNTEC-2000 areal detector. The X-ray source (Cu-Kα) consisted of a tuned monochromatic and parallel X-ray beam (accelerating voltage: 45 kV, tube current: 0.650 mA). The samples were measured on reflection mode in four frames with θ1= θ2 (180 s per frame) and a step size of 2θ = 23° (first frame θ = 12°)." Electrochemical characterization in 0.5 M H 2 SO 4 The characterization protocol for the RDE measurements of all catalysts is summarized in Table  S1. wt.% of the ionomer was sprayed directly to the Aquivion (Solvay™). For more details the reader is referred to reference. [1] The measurements were performed in the chamber of NAP-XP spectrometer at 25°C under 3 mbar oxygen-free water vapor ambient. The MEA resistance determined by high frequency impedance spectroscopy before and after the NAP-XPS measurements was equal to 30 Ohm. This value was used to perform the Ohmic drop (iR) correction. NAP-XPS measurements were performed under constant voltage applied between the WE and the CE. The spectroscopic measurements were performed after the stabilization of the current values (ca. 2-3 min after the potential application) Current transients at selected voltage values are represented in Figure S1.

Reference binding energy (BE) values for Ir species
Literature BE values for various Ir species are shown in Table S1. It should be noted that due to final state effects Ir(III) is characterized by higher BE values BE values compared to Ir(IV) species (see Table S1 and references therein). Rutile-type IrO 2 oxide powder was prepared by thermal oxidation of Ir nanoparticles at 490°C under the air and used as a reference (see Figure S2). The fitting procedure for IrO 2 -rutile was based on the method reported elsewhere. [2] The analysis of the XP spectra revealed some differences between the rutile-type IrO 2 and electrochemical oxide formed on the surface of Ir nano catalyst in terms of the BE (61.8 eV vs. 62.3 eV) and the full width at half maximum (1.0 eV vs. 1.3 eV) suggesting that the oxide formed electrochemically on Ir nano particles most likely has an amorphous structure. Note however that XPS is more sensitive to the composition rather than the structure.  [2] IrO 2 (bulk oxide) 60.9 62.1 [6] IrO 2 NPs (thermal oxide) 60.8 61.7 [7] Ir foil Ir electrode (electrochemically oxidized) IrO 2 (thermally oxidized) 61.1 62.9 62.5 [8]

Reproducibility of NAP-XPS measurements
In order to confirm the reproducibility of the obtained results, the NAP-XPS measurement protocol was applied to a second MEA containing a similar Ir-nano 99.8 electrocatalyst at the anode. The presence of metallic Ir, Ir III and Ir IV components was observed and their potential dependence is shown in Figure S3, reflecting the same trends as those discussed in the main text of the manuscript.

NAP-XPS measurements. Depth profiling
One of the advantages of the synchrotron radiation is the ability to tune the photon energy allowing one to vary the depth of the analyzed sub-surface region. The incident photon energies used in this work are: 460, 595 and 1080 eV, which correspond to 1.9, 2.3 and 3.4 nm depth (estimated as three times the inelastic mean free path), respectively. The contributions of metallic Ir, Ir III and Ir IV determined with various photon energies are shown in Figures S4A-C as a function of the applied voltage. One may see that the contributions of the three components show little dependence of the photon energy. Such a behavior does not support a core-shell morphology but rather suggests that oxidation of Ir nanoparticles results in an inhomogeneous porous (hydr)oxide layer.

Simulation of Ir4f XP spectra
To estimate the thickness of the oxide shell on the surface of electrochemically oxidized Ir nanoparticles the SESSA software was utilized. Figure S5 shows the morphology used for the simulations.

MEA tests in PEM electrolyzer
PEM electrolyzer constant operation tests at 1 and 2 A cm -2 , with MEAs having Ir black (Umicore) and Ir-nano 99.8 anodes were performed. The evolution of cell voltage with respect to time is presented in Figure S7. One can observe that the cell potential of Ir-nano 99.8 is slightly lower than the one with Ir-black, which can be attributed to the improved OER activity of the synthesized catalyst. However, the enhancement in activity cannot be one-to-one correlated with the RDE results presented in Figure 5a of the main text. The MEA manufacture greatly depends on the coating technique, ionomer content, hot-pressing pressure time/temperature, and numerous other engineering parameters. Therefore, it cannot be expected that these parameters should be the same for manufacturing MEAs with catalysts having large differences in OER activity between them. Yet, the PEM electrolyzer operation in both cases is constant during the measured time scale, even after spontaneous PEM electrolyzer system shutdowns. These initial durability tests demonstrate that the MEA with synthesized Ir-nano 99.8 at least is not less stable than the benchmark MEA commercial Ir-black catalyst. [9] Longer duration tests with high performance MEAs are imperatively necessary for commercial applications of the catalyst developed in this work. Figure S7. Two-cell stack with Ir-black (1 mg Ir cm -2 ) and Ir-nano 99.8 (1 mg Ir cm -2 ) with 25 cm² active cell area operating at 1 A cm -2 and 2 A cm -2 , 80 °C, 1 bar for 100 h. Nafion 212 was used as PEM.