Role of boron and phosphorus in enhanced electrocatalytic oxygen evolution by nickel borides and nickel phosphides

The authors acknowledge support by the BMBF in the framework of the project “NEMEZU” (03SF0497B). SMS acknowledges funding from “Programa Internacional de Becas ”la Caixa“‐Severo Ochoa”. J.A. and S.M.S. acknowledge funding from Generalitat de Catalunya 2017 SGR 327 and the Spanish MINECO project VALPEC (ENE2017‐85087‐C3). ICN2 acknowledges support from the Severo Ochoa Programme (MINECO, Grant no. SEV‐2013‐0295) and is funded by the CERCA Programme/Generalitat de Catalunya. Part of the present work has been performed in the framework of Universitat Autonoma de Barcelona Materials Science PhD program.


TOC GRAPHIC
The turn of the 21 st century has witnessed unprecedented pursuit of environmentally friendly and sustainable energy technologies owing to the depletion and eco-unfriendliness of the presentday energy systems and their associated contribution to global warming and climate change. (1) Renewable power-to-hydrogen production by electrochemical water splitting is one of the most promising means for storing and enhancing the energy obtained from renewable energy sources. (2) The low efficiency of the energy recovery cycle due to inefficient electrocatalysis in contemporary regenerative power-to-hydrogen, and hydrogen-to-power energy systems, coupled with the high cost of the systems due to their reliance on the costly platinum group metals (PGM) as electrocatalysts have impeded their entry to a wider market. (3) To this end, reducing the use of the PGM, or better, completely replacing them with efficient low-cost electrocatalysts can tremendously improve the competitiveness of power-to-hydrogen and hydrogen-to-power renewable energy systems.
Compounds and alloys of Co, Ni and Fe, among others, with the pnictogens, chalcogens and some metalloids, (4)(5)(6)(7)(8) have tremendous bonding and structural diversity (9,10) and have recently emerged as very promising low-cost catalysts for the oxygen evolution reaction (OER) in alkaline electrolytes. Studies on these catalysts have in the past, however, focused mainly on enhancing activity through optimization of syntheses, meanwhile, the origin of enhanced electrocatalytic O2 evolution remains unsatisfactorily explained. Additionally, the dependence of the OER activity on alloy stoichiometry and crystal structure has been reported but not adequately explained. Understanding the origin of activity enhancement in these compounds is an important step to further improve their activity and use them as models for the development of new advanced catalysts.
In this work, we present insight into the role of B and P in enhancing OER electrocatalysis by nickel, as observed in nickel phosphides and nickel borides. In nickel borides, electrons are drawn from boron towards nickel, whereas, in nickel phosphides, net electron transfer is towards phosphorus, rendering electronic structure modification inadequate to explain the origin of the OER activity enhancement. Substantially lower apparent energies of OER activation were observed when using Ni2P and Ni2B as OER catalysts compared to pure Ni nanoparticles. The geometric changes induced by B and P on the lattice structure of Ni, notably, the Ni-Ni bond distances, as well as the Ni-P and Ni-B bond distances, thus appear to be the decisive factors for enhanced OER activity in Ni2B and Ni2P.
We used semi-crystalline Ni2B and commercial Ni2P as model catalysts to investigate the role of B and P in promoting electrocatalytic oxygen evolution in nickel borides and nickel phosphides. Ni2B was prepared according to a previously reported procedure (7), as described in the supporting information, while Ni2P was a commercial sample from Sigma Aldrich. XRD analysis of Ni2P indicates predominance of the Ni2P phase ( Fig S1), with minor impurities of Ni3P and Ni12P5. In studying the electrocatalytic behavior of these materials, it is important to point out an often overlooked or not explicitly disclosed fact that these kind of materials tend to undergo tremendous transformation under electrochemical polarization depending on their initial state. (7) For example, the intensity of the Ni 2+  Ni 3+ oxidation peak of Ni2B increased with potential cycling up to a ~5-fold factor before a steady state was reached (Fig.1a). Increase in the intensity of the Ni 2+  Ni 3+ oxidation peak indicates growth of a NiOOH layer and is often accompanied with increase of the OER activity. This observation stresses the necessity to apply a suitable conditioning step to the catalyst films to bring them to a steady state in order to enable accurate determination of their activity and thus a fair comparison. For all the catalysts reported in this study, 100 conditioning CVs were applied to the electrodes prior to recording activity measurements. Fig. 1b   The influence of the Ni:B/P stoichiometry on the OER is illustrated in Fig. 1d by comparison of the OER activity of Ni2B and Ni3B. The synthesis and characterization of Ni3B was reported previously. (7) The Ni:B ratio apparently affects the potential for Ni 2+  Ni 3+ oxidation, as indicated by a higher potential of the Ni 2+  Ni 3+ oxidation peak. Both catalysts begin to evolve oxygen at about the same potential, however, the current rises more sharply in the case of Ni2B owing to a lower Tafel slope (57.8 mV/dec) compared to that of Ni3B (62.5 mV /dec). The OER is therefore not only affected by the presence of B but also by the Ni:B ratio in the catalyst. A similar effect has been reported for nickel phosphide based hydrogen evolution catalysts. (11) The inability to reliably determine the TOF of the catalysts owing to non-participation of all available Ni atoms in the reaction implies that the TOF is not a decisive metric for describing the intrinsic activity of the catalysts.
The activation energy of a reaction is an intensive quantity that only depends on the inherent properties of the catalyst and is insensitive to extrinsic factors that accrue from the interaction of the electrode with the electrolyte, and to extended morphological features of the surface. We recorded the OER on Ni2B, Ni2P and Ni at different electrolyte temperatures from 20 o C to 65 o C. Fig. 2a shows LSVs recorded on a Ni2P modified electrode at different electrolyte temperatures. As can be seen, the rate of O2 evolution increased with temperature as manifested by decrease of the OER overpotential with temperature and higher electrocatalytic currents. This dependence implies that the reaction can be treated by Arrhenius kinetics to determine the activation energy of the OER on the various catalysts. However, since the OER is a multi-step reaction involving many intermediates and activation steps, with the rate-limiting step not known, we have chosen to refer to the determined activation energy as apparent energy of activation, denoted as Ea*.  , by plotting log i versus 1/T. W decreased linearly with  as shown in Fig. 2b. The apparent energy of activation (Ea*) at Eeq was obtained from the intercept of the graph at  = 0, and was 78.4 kJ mol -1 for Ni2P, 65.4 kJ mol -1 for Ni2B and 94.0 kJ mol -1 for Ni nanoparticles. A study by Bowden, (12) showed that for a given reaction, small surface impurities can significantly affect the value of  and Ea*. Since Ni2B, Ni2P and Ni were investigated under similar conditions, these values represent inherent properties of the respective catalysts. The guest elements (B and P) therefore obviously affect the intrinsic electrocatalytic activity of the host element (Ni).
We probed the influence of B and P on the surface electronic properties of Ni by means of XPS. Fig. 3a shows high-resolution XPS spectra of the Ni 2p region of Ni2P and Ni2B. The binding energy (BE) of the Ni 2p3/2 peak was 852.52 eV for Ni2B and 852.30 eV for Ni2P, translating into chemical shifts of -0.08 eV and 0.40 eV respectively, with respect to the BE of the 2p3/2 peak at 852.60 eV for pure nickel. (13) These results are consistent with the work of Okamoto et al. and others, who reported electron enrichment of Ni in the presence of boron, whereas P accepted electrons from Ni. (14)(15)(16)(17) Therefore, the fact that B and P induce two opposite electronic effects on Ni yet both lead to enhanced electrocatalysis of the OER Ni cannot be rationalized by electronic structure modification. We recognize that there is no unified theory explaining electron transfer in metal-metalloid compounds and metallic glassy alloys and some controversy exists, (18,19) so our inferences are confined to the experimental observations. A discuss of electron transfer theories in these compounds is outside the scope of this work. Importantly, XPS post-mortem analysis of both Ni2B and Ni2P after electrochemical activation indicates that their surfaces are similar and covered with NiOOH. (7) Upon polarization, both Ni2B and Ni2P adopt a core-shell structure, where the surface is composed of NiOOH while the core is Ni2P and Ni2B. This was confirmed by XPS depth profiling by Ar + sputtering, which revealed increase in the Ni 0 /Ni 2+ ratio with sputtering time, and a clear increase in the intensity of the Ni 2p3/2 peak at 852.52 eV, as illustrated in Fig. 3b for the case of activated Ni2B.
Evidence of the core-shell structure of the activated Ni2P catalyst is provided by the elemental intensity maps showing the spatial distribution of Ni, P and O (Fig. 3c), obtained by Electron Energy Loss Spectroscopy (EELS). As we can observe, Ni and P are mainly concentrated in the bulk of the particle meanwhile oxygen is on the surface. In addition, high resolution transmission electron microscopy analysis of the particles (not shown here, see SI) and the corresponding obtained power spectra (FFT) as the example shown in Fig 3c and detailed studies in SI, revealed that they have a NiP orthorhombic structure with Pbca space group.
Essentially, the modification of Ni with B or P leads to expansion of the Ni-Ni bond distance, and various Ni:B/Ni:P ratios and crystalline structures give rise to dissimilar Ni-Ni distances. As an illustration, the unit cell structures of Ni3B, Ni2B and NiB are presented in Fig. 3 to show the effect of the Ni:B stoichiometry and crystal structure on the Ni-Ni bond distances. For example, the nearest Ni-Ni bond distance in Ni2P is 2.640 Å whereas it is 2.563 Å in Ni5P4, both longer than 2.492 Å, the Ni-Ni bond length in pure Ni. (20) As shown in Fig. 1d, variation of the Ni:B/Ni:P ratio influences the OER activity. Higher complexity in underpinning the role of B and P arises if besides variation in the lattice parameters, subtle changes in the electronic structure of the host metal accompany variation of the Ni:B/Ni:P stoichiometry. We previously observed that under OER conditions, surface B and P get oxidized to form their respective oxo-species, borates and phosphates. (6,7) These species being anionic obviously introduce unique interactions at the electrode/electrolyte interface, which are inexistent on pure Ni electrodes. Therefore, in addition to deciphering the complex interplay between the electronic and geometric factors, solid understanding of the nature of electrode/electrolyte interactions, specifically, how the B and P oxo-species contribute to these interactions, and the ultimate effect on the mechanism and kinetics of the OER is imperative in order to get a holistic interpretation of the role of B and P in promoting the OER activity of Ni.
In conclusion, in nickel borides and phosphides, the guest atoms (boron and phosphorus) induce opposite electronic effects on the host atom (nickel), yet both elements bring about enhanced electrocatalysis of oxygen evolution. The origin of OER activity enhancement cannot therefore be rationalized by electronic structure modification. The presence of B or P leads to a decrease of the energy of activation of the OER, indicating intrinsic changes in the electrocatalytic properties of nickel. Changes in the lattice structure induced by the presence of B or P, specifically, the atomic order of the Ni atoms, and the Ni-Ni bond distances, therefore seem to be the dominant factors for electrocatalytic enhancement of the OER in nickel borides and nickel phosphides.

Catalyst synthesis
Nickel phosphide (Ni2P) powder (~100 mesh, 98%) and nickel nanopowder were from Sigma Aldrich. Semi-crystalline Ni2B was prepared by reaction of a solution of NiCl2 (1.0 M) with NaBH4 (1.0 M) in NaOH (1.0) followed by annealing at 300 o C under argon for 2 hours. (7) XRD and XPS characterization XRD measurements were performed using a PANalytical theta-theta powder diffractometer equipped with a Cu-K -radiation source.
XPS measurements were carried out in an ultra-high vacuum set-up (UHV) equipped with a high resolution Gammadata-Scienta SES 2002 analyzer. A monochromatic Al Kα X-ray source

Transmission Electron Microscopy
High-resolution Transmission Electron Microscopy (HRTEM) and scanning TEM (STEM) studies were carried out using a field emission gun FEI Tecnai F20 microscope at 200 kV with a point-to-point resolution of 0.19 nm. High angle annular dark-field (HAADF) STEM was