Plasma-Assisted Synthesis of NiCoP for Efficient Overall Water

Nov 9, 2016 - Efficient water splitting requires highly active, earth-abundant, and robust catalysts. Monometallic phosphides such as Ni2P have been s...
5 downloads 6 Views 8MB Size
Download PDF
Letter pubs.acs.org/NanoLett

Plasma-Assisted Synthesis of NiCoP for Efficient Overall Water Splitting Hanfeng Liang, Appala N. Gandi, Dalaver H. Anjum, Xianbin Wang, Udo Schwingenschlögl, and Husam N. Alshareef* Materials Science and Engineering, King Abdullah University of Science and Technology (KAUST), Thuwal23955-6900, Saudi Arabia S Supporting Information *

ABSTRACT: Efficient water splitting requires highly active, earth-abundant, and robust catalysts. Monometallic phosphides such as Ni2P have been shown to be active toward water splitting. Our theoretical analysis has suggested that their performance can be further enhanced by substitution with extrinsic metals, though very little work has been conducted in this area. Here we present for the first time a novel PH3 plasma-assisted approach to convert NiCo hydroxides into ternary NiCoP. The obtained NiCoP nanostructure supported on Ni foam shows superior catalytic activity toward the hydrogen evolution reaction (HER) with a low overpotential of 32 mV at −10 mA cm−2 in alkaline media. Moreover, it is also capable of catalyzing the oxygen evolution reaction (OER) with high efficiency though the real active sites are surface oxides in situ formed during the catalysis. Specifically, a current density of 10 mA cm−2 is achieved at overpotential of 280 mV. These overpotentials are among the best reported values for non-noble metal catalysts. Most importantly, when used as both the cathode and anode for overall water splitting, a current density of 10 mA cm−2 is achieved at a cell voltage as low as 1.58 V, making NiCoP among the most efficient earth-abundant catalysts for water splitting. Moreover, our new synthetic approach can serve as a versatile route to synthesize various bimetallic or even more complex phosphides for various applications. KEYWORDS: NiCoP, plasma synthesis, DFT, HER, OER, electrocatalysis

M

bind to the catalyst to produce H2.19,21,22 Ni2P has shown the best activity toward the HDS among the monometallic phosphides and other compounds; recent efforts have been directed toward further enhancing its performance by substituting other transition metals such as Mo and Co.23−25 This is because a synergistic effect between these components was foreseen as already confirmed for bimetallic sulfides. Unexpectedly, however, the bimetallic phosphides did not show superior activity to the monometallic phosphides expect for the case of NixCoyP where a 50% increase in conversion was achieved.26 Given the commonalities between the HDS and HER, we predicted that the Co substitution would also enhance the HER activity of Ni2P. We noticed that it has been previously demonstrated by Jin et al. that the substitution of one S atom of CoS2 with P atom could alter the electronic structure and consequently dramatically enhance the HER performance.27 In fact, the resultant CoPS is currently the best non-noble metal HER catalyst in acidic media. We hence carried out density function theory (DFT) calculations to determine if the Co substitution influences the electronic structure of Ni2P and the surface adsorption energy of the reactants and thus the HER activity. Both the Ni2P and NiCoP adopt the hexagonal Fe2P structure (see Figure 1a,b, and also

olecular hydrogen (H2) is a clean and sustainable energy carrier that has the potential to meet the ever-growing global energy demands at no environmental cost.1,2 Water splitting is one of the most promising ways to produce hydrogen but it is a thermodynamically uphill reaction and thus requires external energies such as electricity to initiate the reaction. The water electrolysis can be promoted by highly efficient and robust catalysts that can substantially expedite the sluggish kinetics of the two half reactions, namely the hydrogen evolution and the oxygen evolution reactions (HER and OER).3−6 Precious metal-containing compounds are at present the state-of-the-art catalysts (Pt for HER, whereas RuO2 and IrO2 for OER) but their large-scale application is seriously limited by the high cost and scarcity. Motivated by this challenge, enormous efforts have been devoted to developing cost-efficient alternatives including sulfides, selenides, phosphides, and many other nonprecious transition metal compounds.3−6 Among them, metal phosphides such as Ni2P, CoP, MoP, and WP have recently attracted extensive research interest owing to their high activity.7−16 These compounds are wellknown hydrodesulfurization (HDS) catalysts and recently have demonstrated their efficacy for HER catalysis, which is not surprising because both the HDS and HER rely on the reversible adsorption and desorption of hydrogen on the catalyst surface.15,17−20 In HDS, the hydrogen dissociates and then reacts with sulfur to form H2S, whereas in HER, protons © XXXX American Chemical Society

Received: September 9, 2016 Revised: October 19, 2016

A

DOI: 10.1021/acs.nanolett.6b03803 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Figure 1. Crystal structures and theoretical predictions of the properties of Ni2P and NiCoP. (a,b) Crystal structures in unit cell of Ni2P and NiCoP. (c,d) Electronic band structures and projected densities of states of the bulk Ni2P and NiCoP (spin-up, green curves; spin-down, red curves), respectively. (e) Free-energy diagram for H adsorption on the Ni2P, NiCoP (0001) and Pt(111) surfaces. (f) Adsorption energy for H2O.

Figure 2. Synthetic route and structural characterization of the NiCoP nanostructure. (a) Schematic illustration of the synthetic route for NiCoP nanostructure on Ni foam. (b) PXRD patterns of NiCo−OH and the converted NiCoP. The asterisks mark the diffraction peaks from Ni foam. (c) SEM image of NiCo−OH. (d) SEM images and (e) corresponding EDS elemental maps of the NiCoP. (f,g) TEM images, (h) SAED pattern, and (i) high-resolution TEM image and EDS spectrum (inset) of the NiCoP. The dashed white line highlights the crystalline−amorphous boundary. B

DOI: 10.1021/acs.nanolett.6b03803 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Figure 3. XPS characterization of the NiCo−OH precursor and the NiCoP. (a) Ni 2p3/2 spectra, and (b) Co 2p3/2 spectra of the NiCo−OH precursor and the NiCoP. (c) P 2p spectrum of the NiCoP. Circles, raw data; gray lines, background; yellow lines, overall fit; and other colored lines, fit components.

between Ni and Co during OER,32,33 as well as the metallic nature and strong ability of water adsorption on NiCoP, one can expect that the NiCoP could potentially serve as an efficient OER catalyst (through the real active sites for OER are metal oxides as demonstrated later). Motivated by these promising predictions, we set out to design and synthesize a bimetallic phosphide phase, namely NiCoP, and further establish it as a cost-efficient and highly active electrocatalyst for both the HER and OER in alkaline media. The NiCoP supported on Ni foam (NiCoP/NF) was synthesized using a novel PH3 plasmaassisted process and the resulting electrodes achieved stable performance for both the HER and OER that are superior to most of the catalysts reported so far. Furthermore, the NiCoP/ NF is also capable of catalyzing overall water splitting with a cell voltage as low as 1.58 V at the current density of 10 mA cm−2 and outstanding long-term durability. The NiCoP nanostructure supported on Ni foam was converted from hydrothermally formed NiCo hydroxide (NiCo−OH) precursor using a novel PH3 plasma-assisted process as illustrated in Figure 2a (see Supporting Information for details). Compared to conventional synthetic methods for phosphides,13 our newly developed plasma-assisted process enables low temperature, fast preparation and could be, in principle, applied to synthesize various metal phosphides. Most importantly, because of the high reactivity of plasma we can even directly convert metals such as Cu into Cu3P using this unique technique at a relatively low temperature (Figure S4), which is otherwise much more difficult to synthesize through commonly practiced methods. Note that for the synthesis of NiCoP, the temperature was set at 250 °C to avoid the phosphorization of Ni foam, which occurs at 300 °C (Figure S4). We first characterized the crystal structure and the morphology of the NiCoP. The powder X-ray diffraction (PXRD) patterns (Figure 2b) clearly suggest that the NiCo− OH precursor was converted into hexagonal NiCoP (space group P6̅2m, a = 5.834 Å), further confirming the efficacy of converting hydroxides to phosphides by PH3 plasma treatment. The scanning electron microscopy (SEM) image shows that the NiCo−OH (Figure 2c) is composed of many nanoplates, a typical morphology for hydroxides. These nanoplates are grown uniformly on the Ni foam and are interconnected with each other, forming a wall-like structure. This morphology is largely

Figure S2 in Supporting Information). As demonstrated in Figure 1c,d, Ni2P and NiCoP show no band gap, that is, a metallic nature, which would favor the electron transfer in catalysis. The d-states of NiCoP are found closer to the Fermi level as compared to Ni2P, which points to a lower intermedia adsorption energy. The HER pathway in acidic media is generally depicted as three-state diagram with initial state of H+ + e, intermediate adsorbed H* (* being the adsorption site), and product of 1/2H2.28,29 The Gibbs free energy of the adsorbed H*, |ΔGH*|, is commonly used to evaluate the HER activity of catalysts.28,29 We thus calculated the Gibbs free energy for H adsorption (ΔGH*) on the (0001) surface of Ni2P and NiCoP (see Supporting Information for calculation details). Because the HER relies on the reversible adsorption and desorption of H, ideally the H binding should be neither too strong nor too weak, that is, |ΔGH*| should be close to thermo-neutral (|ΔGH*| = 0).29 According to Figure 1e, ΔGH* of the Ni2P (0001) surface is −0.86 and −0.45 eV for sites 1 and 2, respectively, see the adsorption sites in Figure S3. Co substitution lowers the ΔGH* on the NiCoP (0001) surface dramatically to −0.15 (site 1) and −0.14 eV (site 2), which are close to the value on the Pt(111) surface. Note that the HER in alkaline media requires an additional first step of water dissociation;3 we thus further calculated the adsorption energy of an H2O molecule (ΔEH2O) on the Ni2P and NiCoP (0001) surfaces and found in both cases values (absolute value) of more than 18 eV, indicating strong ability of water adsorption (Figure 1f). In fact, the adsorption of water is much more favorable on NiCoP (0001) surface than that on Pt(111) surface, which possesses an absolute adsorption energy of 8.97 eV. Therefore, the DFT result suggests that the NiCoP is indeed a highly active HER catalyst in alkaline media. Besides the HER, the other half reaction of water splitting, namely OER, is also of key importance. In fact, the OER has long been the bottleneck because of its sluggish kinetics. Efficient water splitting requires the optimization of both the HER and OER catalysts.3 Ideally, the water splitting should be driven by the same catalyst, which would greatly simplify the water splitting system and make water electrolysis feasible at the industrial scale. Interestingly, some monometallic phosphides such as Ni2P and CoP have recently been reported to act as OER catalysts.7,8,30,31 Given the synergistic effect C

DOI: 10.1021/acs.nanolett.6b03803 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Figure 4. Hydrogen evolution reaction electrocatalysis in 1 M KOH. (a) IR-corrected polarization curves per geometric area of the NiCoP/NF recorded at a scan rate of 3 mV s−1, along with Ni2P/NF, NiCo−OH/NF, and NF for comparison. (b) Polarization curves-derived Tafel slopes for the corresponding electrocatalysts. (c) Long-term stability test carried out under a constant current density of −10 mA cm−2. (d) Difference in current density plotted against scan rate showing the extraction of the double-layer capacitances. (e) Polarization curves from (a) normalized to the electrochemical active surface area (ECSA). (f) The H2 turn over frequencies (TOFs) per surface site over different catalysts.

retained after the phosphorization. As shown in Figure 2d, the NiCoP exhibits a hierarchical structure with thin nanoplates lying aslant or perpendicular to the substrate. Such highly open network could possibly have good mechanical strength and enable a close contact with the electrolyte, which is believed to be beneficial for electrocatalysis.32,34 The energy dispersive Xray spectroscopy (EDS) mapping (Figure 2e) confirms the presence of Ni, Co, and P elements. The ratio of Ni/Co/P was determined to be 1.106:1:1.138, giving a stoichiometric formula of Ni1.11CoP1.14, close to that of NiCoP. The atomic distribution of Ni, Co, and P is further revealed by the scanning transmission electron microscopy-electron energy loss spectroscopy (STEM-EELS) elemental mapping (Figure S5). Though the overall morphology barely changes, the TEM image reveals that many pores are distributed uniformly throughout the NiCoP (Figure 2f) upon phosphorization. A close observation (Figure 2g) further shows the NiCoP plate is composed of numerous small nanocrystallites. The polycrystalline nature is further evidenced by the corresponding selected area electron diffraction (SAED) pattern (Figure 2h), which can be indexed into hexagonal NiCoP, consistent with the PXRD result. We then performed high-resolution TEM (HRTEM) to examine the nanostructure in more detail. As shown in Figure 2i, small crystalline NiCoP nanoparticles are surrounded by amorphous-like shells (∼2 nm), suggesting the surface of NiCoP may be oxidized to form phosphates due to air contact. This explains the high amount of O detected by EDS (inset of Figure 2i; the Cu and C come from TEM grid). Such crystalline−amorphous structure could protect the inner NiCoP from being etched to some degree during the catalysis and more importantly would promote the electron transfer from the metallic inner NiCoP to the outer shell and thus improve the catalytic activity. For comparison, we also

synthesized Ni2P with similar platelike porous nanostructure using Ni(OH)2 as precursor. The synthesis and detailed structural characterization (Figure S6) can be found in Supporting Information. We further carried out the X-ray photoelectron spectroscopy (XPS) measurements to probe the surface composition and the oxidation state of the NiCoP. The Ni 2p3/2 core level spectrum (Figure 3a) of NiCo−OH shows two main peaks at binding energies of 855.6 and 861.1 eV that can be assigned to Ni(OH)x and its satellite peak.35 After phosphorization, three peaks with binding energies of 853.0, 856.3, and 861.1 eV are observed, which should be related to the Ni−P, Ni−POx, and the satellite peak, respectively. Note that the binding energy of 853.0 is very close to that of metallic Ni (852.6 eV),36 suggesting the presence of partially charged Ni species (Niδ+, δ is likely close to 0). Similarly, the Co 2p3/2 segment of the converted NiCoP (Figure 3b) possesses a new peak located at 778.5 eV as a consequence of the formation of Co−P.20,31 This binding energy is also found to be slightly higher than that of metallic Co (778.2 eV),37 indicating that the Co carries a partially positive charge (Coδ+). The peak at 781.5 eV could be ascribed to a Co oxidation state, which is associated with Co− POx.31Figure 3c shows the P 2p region, where two doublets with main peak binding energies at 129.1 and 132.9 eV are observed. The former can be assigned to reduced phosphorus in the form of metal phosphides, and the latter to phosphate species (P5+).31,38 The binding energy of 129.3 eV is slightly lower than that of elemental P (130.0 eV),39 which suggests the P is partially negatively charged (Pδ‑). Thus, the P can act as base to trap positively charged protons during electrocatalysis. This feature is similar to those observed in [NiFe] hydrogenase,19 which is a highly active biological HER catalyst. The XPS result further confirms the formation of NiCoP with D

DOI: 10.1021/acs.nanolett.6b03803 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Figure 5. Oxygen evolution reaction electrocatalysis in 1 M KOH. (a) IR-corrected polarization curves per geometric area of the NiCo-P/NF recorded at a low scan rate of 0.5 mV s−1, along with Ni−P/NF, NiCo−OH/NF, and NF for comparison. (b) Polarization curves-derived Tafel slopes for the corresponding electrocatalysts. (c) Long-term stability test carried out at a constant current density of 10 mA cm−2. (d) Plots showing the extraction of the double-layer capacitances allows the estimation of the electrochemically active surface area. (e) Polarization curves from (a) normalized to the ECSA. (f) The O2 turn over frequencies (TOFs) per surface site over different catalysts.

mapping suggests that this thin layer consists of Ni, Co, and O elements, while no P was detected. XPS (Figure S10) on the NiCoP after HER further reveals the disappearance of the low energy peaks of Ni (853.0 eV), Co (778.5 eV), and P (129.3 eV). These observations indicate the surface of the NiCoP is dominated by NixCoy(OH) after the HER catalysis. Similar phenomenon has been observed for Ni phosphides (Ni5P4/ Ni2P/NiP2), where the formation of amorphous Ni(OH)x was observed after the HER catalysis in alkaline media.41 To gain insights into the intrinsic activity of the catalysts, we measured the double-layer capacitances (Cdl) to estimate the electrocatalytic active surface areas (ECSAs)42,43 and further normalized the geometric current density to the corresponding ECSA (see Supporting Information for details). As shown in Figure 4d, the NiCoP/NF has the biggest Cdl, thus the highest catalytically relevant surface area, which could be mainly attributed to the porous structure and more exposed active sites. When normalized to the ECSA (Figure 4e), the NiCoP/ NF displays the highest current density, followed by Ni2P and then NiCo−OH, indicating the NiCoP is intrinsically more active than Ni2P. We also calculated the H2 turn over frequencies (TOFs) per surface site for these catalysts to further evaluate their intrinsic HER activity. Here we assume all surface sites (including both the transition metal and P atoms) are involved in the HER, which would give a lower limit of the TOF (see Supporting Information for details). Figure 4f shows the calculated TOFs plotted against potential. The average TOF for NiCoP/NF is almost 2 and 13 times larger than those for Ni2P/NF and NiCo−OH/NF. For instance, the experimental TOFs are 8.93, 3.48, and 0.64 s−1 for NiCoP/NF, Ni2P/NF, and NiCo−OH/NF at −100 mV versus RHE, respectively. These observations confirm the predictions from DFT and establish our NiCoP/NF as an outstanding HER

surface partially being oxidized to phosphate as observed from TEM analysis. We also performed XPS analysis on Ni2P (Figure S7). Compared to Ni2P, the Ni 2p3/2 and P 2p peaks of NiCoP slightly shift toward the lower binding energies, indicating a change in the electronic structure occurs due to Co substitution. We then compared the HER performance of the NiCoP, Ni2P and NiCo−OH using a standard three-electrode cell in 1 M KOH (see Supporting Information for details). The comparison in Figure 4a shows that the NiCoP/NF exhibits a much higher activity than Ni2P/NF and NiCo−OH/NF. Specifically, a geometric current density of −10 mA cm−2 can be achieved at overpotential as low as 32 mV on NiCoP/NF (see the original data and the Nyquist plots in Figure S8). In contrast, 93, 201, and 242 mV overpotentials are required for Ni2P/NF, NiCo−OH/NF, and bare NF to drive the same current density, respectively. Note the bare NF only shows negligible current in the potential window of −200 to 0 mV versus the reversible hydrogen electrode (RHE), further supporting the high performance of NiCoP. The NiCoP/NF also shows a smallest Tafel slope of 37 mV dec−1 among the four catalysts investigated (Figure 4b). Such Tafel slope value suggests a two-electron transfer process occurs upon NiCoP/ NF following the Volmer−Tafel mechanism.40 Furthermore, an exchange current density (J0) as high as 1.36 mA cm−2 was obtained for NiCoP/NF based on the intercept of the Tafel plot, which is much larger than that of Ni2P/NF (0.37 mA cm−2). These catalysts were also found highly stable over a 24 h long-term test (Figure 4c). Poststructural characterization (Figure S9) shows that the overall platelike morphology of the NiCoP does not change significantly, and the major phase remains NiCoP as confirmed by the HRTEM. Nevertheless, a thin layer was formed on the surface of the NiCoP plate. EELS E

DOI: 10.1021/acs.nanolett.6b03803 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

reactions. In fact, the real surface active sites of metal phosphides for OER have recently demonstrated to be the metal oxyhydroxides (or oxides) in situ formed during the OER,7,8,30 which are different from those for HER. We thus carried out the XPS measurements to further understand the compositional change after the OER catalysis (Figure S10). After OER catalysis, the low-valent Ni (853.0 eV) and Co (778.5 eV) peaks, which are assigned to Ni−P and Co−P in the as-synthesized sample, disappear, and the P signal becomes very weak and is actually below the detection limit of XPS. This result suggests that the surface-bound phosphates and phosphides are being oxidized to oxides or oxyhydroxides, as commonly observed for metal phosphides after OER catalysis.41 Of note, a P-deficient surface layer can be clearly identified by the STEM-EELS mapping (Figure S12). The resulting NiCoP/NixCoyO (or NixCoyOOH) heterojunction could promote the electron transfer from the metallic NiCoP to the surface layer and thus contribute to the OER activity, as suggested by a previous study on OER over Ni phosphides.41 Encouraged by the high catalytic activity of the NiCoP/NF toward both the HER and OER in alkaline media, we further used the NiCoP/NF as both the anode and cathode for water electrolysis in a single electrolyzer (Figure 6a). Figure 6b shows

catalyst that outperforms almost all the reported earthabundant HER catalysts in alkaline media (see comparison in Table S1) and further proves the efficacy of our strategy to significantly enhance the HER performance of Ni2P by tuning the electronic structure and the hydrogen adsorption energy through Co substitution. We further investigated the OER performance of the NiCoP/NF in 1 M KOH, along with Ni2P/NF and NiCo− OH/NF for comparison (see Supporting Information for details). Figure 5a shows the IR-corrected linear sweep voltammetry (LSV) curves (see uncorrected data and the Nyquist plots in Figure S11). The NiCoP/NF exhibits the highest OER activity and can deliver a geometric current density of 10 mA cm−2 at an overpotential of 280 mV. In contrast, Ni2P/NF and NiCo−OH/NF require overpotentials of 340 and 404 mV to drive the same current density, respectively. Furthermore, the current density obtained for NiCoP/NF at overpotential of 300 mV is 16.5 mA cm−2, which is 3 times larger than that for Ni2P/NF (4.23 mA cm−2). These numbers compare favorably to many other recently reported high-performance OER catalysts (Table S2) and place the NiCoP/NF among the top tier of earth-abundant OER catalysts. All catalysts show a distinct anodic peak between 1.3 and 1.5 V versus RHE, which should be associated with the oxidation of Ni or Co (or both) species.33 For Ni2P/NF, this anodic peak is located at around ∼1.4 V versus RHE, which is in agreement with the reported value for Ni2P/NF.38 Interestingly, such anodic peak observed for NiCoP/NF shifts to lower potential (∼1.35 V versus RHE), suggesting that the electronic structure of Ni2P is likely altered by the presence of Co ions. The Co substitution may lower the thermodynamic barrier of the proton-coupled electron transfer (PCET) preequilibrium and promote the O−O bond formation, thus lowering the activation barrier required for OER, and leading to enhanced catalytic performance.44 The Tafel slopes of these catalysts are found to be very close (Figure 5b), suggesting similar OER kinetics. The stability of these materials was examined by chronopotentiometry and the result shows these catalysts are reasonably stable (Figure 5c). For NiCoP/NF, the potential required to maintain a constant geometric current density of 10 mA cm−2 only slightly increases from 1.51 to 1.53 V versus RHE after 24 h. The overall morphology of the NiCoP plates is also maintained after the OER catalysis (Figure S12). The improved performance of NiCoP might be attributed to the synergistic effect between Ni and Co,32,33 and the increase in ECSA as revealed by the double-layer capacitance measurements (Figure 5d). The Cdl of NiCoP/NF (4.9 mF cm−2) is about 1.5 times as large as that of the Ni2P/NF (3.2 mF cm−2), suggesting more accessible active sites are exposed. The geometric current densities were then normalized to ECSA in order to compare the intrinsic activity of these catalysts. As shown in Figure 5e, the NiCoP/NF again shows the highest normalized current densities at lower overpotential. Furthermore, the calculated TOF (Figure 5f) of the NiCoP/NF at overpotential of 300 mV (1.53 V versus RHE) is 3.88 s−1, more than twice as that of the Ni2P/NF (1.54 s−1). These results confirm the higher intrinsic catalytic activity of NiCoP than Ni2P and the potential of the NiCoP/NF as a cost-efficient and highly active alternative for OER. We noted that the ECSAs of the NiCoP/NF during HER and OER are significantly different, whereas the active sites for HER and OER are not necessarily the same; this is more likely an indication of the surface composition change during the two

Figure 6. NiCoP/NF electrocatalyst for overall water splitting in 1 M KOH. (a) Schematic illustration of two-electrode cell using NiCoP/ NF for both anode and cathode for water splitting. (b) Polarization curve recorded at 0.5 mV s−1. Inset: digital photograph of the twoelectrode configuration. (c) Long-term stability test carried out at constant current densities of 10, 20, and 50 mA cm−2.

the polarization curve recorded at 0.5 mV s−1. Remarkably, the current density of 10 mA cm−2 can be achieved at a cell voltage as low as 1.58 V, and 100 and 200 mA cm−2 at 1.82 and 1.98 V, respectively. These numbers are better than or at least comparable to previously reported transition metal phosphide, chalcogenide, and hydroxide-based electrocatalysts (Table S3). Furthermore, the NiCoP/NF shows very good stability upon long-term testing and only a slight deactivation is observed after 24 h (Figure 6c). The excellent catalytic activity of the NiCoP/NF could be attributed to the following factors: (i) The Co substitution F

DOI: 10.1021/acs.nanolett.6b03803 Nano Lett. XXXX, XXX, XXX−XXX

Nano Letters

■

alters the electronic structure and further optimizes the reversible adsorption and desorption of hydrogen (Figure 1c). (ii) The metallic nature of the NiCoP/NF ensures good electric conductivity thus favors fast electron transport (Figure 1b). For comparison, the annealed product (see Supporting Information for synthetic details), namely NiCo oxides/NF, shows inferior performance for both the HER and OER (Figure S13), which is likely due to the poor intrinsic electric conductivity of oxides. (iii) The fast synthesized NiCoP nanocrystals were found rich of defects (Figure S14), which promote the electrocatalytic activity.45 (iv) The nanopores generate abundant active sites, leading to improved catalytic activity per geometric area. (v) The porous structure enables close contact with electrolyte thus allows more efficient utilization of active site. (vi) The highly open hierarchical structure facilitates the gas release during the HER and OER catalysis. (vii) The direct integration of NiCoP onto Ni foam not only enables good mechanical adhesion and electric connection but also avoids the use of extra binders. (vii) The Ni foam support also contributes to the catalysis to some degree. In conclusion, we present a combined theoretical and experimental study to establish NiCoP as a highly active and earth-abundant electrocatalyst for water splitting. The NiCoP was synthesized using a novel PH3 plasma-assisted method, which in principle could serve as a versatile route to synthesize various bimetallic or even more complex phosphides. The assynthesized NiCoP porous nanostructures supported on Ni foam achieved superior performance for both the HER and OER (note that the real active sites for OER are NiCoP-derived oxides) in alkaline media than reported metal phosphides and are among the most active earth-abundant electrocatalysts developed so far. Our result suggests that substituting a second metal into monometallic phosphides could effectively alter the electronic structure of the parent compounds and further tune the hydrogen (or water) adsorption energy, thus dramatically enhance the catalytic activity. This strategy might be generally applicable for enhancing the electrocatalytic and many other applications (e.g., batteries, supercapacitors) of transition metal phosphides.

■

REFERENCES

(1) Lewis, N. S.; Nocera, D. G. Proc. Natl. Acad. Sci. U. S. A. 2006, 103 (43), 15729−15735. (2) Nocera, D. G. Acc. Chem. Res. 2012, 45 (5), 767−776. (3) Jiao, Y.; Zheng, Y.; Jaroniec, M.; Qiao, S. Z. Chem. Soc. Rev. 2015, 44 (8), 2060−2086. (4) Zheng, Y.; Jiao, Y.; Jaroniec, M.; Qiao, S. Z. Angew. Chem., Int. Ed. 2015, 54 (1), 52−65. (5) Faber, M. S.; Jin, S. Energy Environ. Sci. 2014, 7 (11), 3519−3542. (6) Hong, W. T.; Risch, M.; Stoerzinger, K. A.; Grimaud, A.; Suntivich, J.; Shao-Horn, Y. Energy Environ. Sci. 2015, 8 (5), 1404− 1427. (7) Stern, L.-A.; Feng, L.; Song, F.; Hu, X. Energy Environ. Sci. 2015, 8 (8), 2347−2351. (8) Wang, P.; Song, F.; Amal, R.; Ng, Y. H.; Hu, X. ChemSusChem 2016, 9, 472−477. (9) Yang, H.; Zhang, Y.; Hu, F.; Wang, Q. Nano Lett. 2015, 15 (11), 7616−7620. (10) Liu, Q.; Tian, J.; Cui, W.; Jiang, P.; Cheng, N.; Asiri, A. M.; Sun, X. Angew. Chem. 2014, 126 (26), 6828−6832. (11) Xiao, P.; Sk, M. A.; Thia, L.; Ge, X.; Lim, R. J.; Wang, J.-Y.; Lim, K. H.; Wang, X. Energy Environ. Sci. 2014, 7 (8), 2624−2629. (12) Xing, Z.; Liu, Q.; Asiri, A. M.; Sun, X. ACS Catal. 2015, 5 (1), 145−149. (13) Shi, Y.; Zhang, B. Chem. Soc. Rev. 2016, 45 (6), 1529−1541. (14) Xiao, P.; Chen, W.; Wang, X. Adv. Energy Mater. 2015, 5, 1500985. (15) Popczun, E. J.; McKone, J. R.; Read, C. G.; Biacchi, A. J.; Wiltrout, A. M.; Lewis, N. S.; Schaak, R. E. J. Am. Chem. Soc. 2013, 135, 9267−9270. (16) Callejas, J. F.; Read, C. G.; Roske, C. W.; Lewis, N. S.; Schaak, R. E. Chem. Mater. 2016, 28, 6017−6044. (17) Tian, S.; Li, X.; Wang, A.; Prins, R.; Chen, Y.; Hu, Y. Angew. Chem., Int. Ed. 2016, 55, 4030−4034. (18) Zhao, H.; Oyama, S. T.; Freund, H.-J.; Włodarczyk, R.; Sierka, M. Appl. Catal., B 2015, 164, 204−216. (19) Liu, P.; Rodriguez, J. A. J. Am. Chem. Soc. 2005, 127, 14871− 14878. (20) Burns, A. W.; Layman, K. A.; Bale, D. H.; Bussell, M. E. Appl. Catal., A 2008, 343, 68−76. (21) Infantes-Molina, A.; Cecilia, J.; Pawelec, B.; Fierro, J.; Rodríguez-Castellón, E.; Jiménez-López, A. Appl. Catal., A 2010, 390, 253−263. (22) Liu, P.; Rodriguez, J. A.; Asakura, T.; Gomes, J.; Nakamura, K. J. Phys. Chem. B 2005, 109, 4575−4583. (23) Rodriguez, J. A.; Kim, J.-Y.; Hanson, J. C.; Sawhill, S. J.; Bussell, M. E. J. Phys. Chem. B 2003, 107, 6276−6285. (24) Zuzaniuk, V.; Prins, R. J. Catal. 2003, 219, 85−96. (25) Abu, I. I.; Smith, K. J. J. Catal. 2006, 241, 356−366. (26) Prins, R.; Bussell, M. E. Catal. Lett. 2012, 142, 1413−1436. (27) Cabán-Acevedo, M.; Stone, M. L.; Schmidt, J.; Thomas, J. G.; Ding, Q.; Chang, H. C.; Tsai, M. L.; He, J. H.; Jin, S. Nat. Mater. 2015, 14, 1245−1251. (28) Greeley, J.; Jaramillo, T. F.; Bonde, J.; Chorkendorff, I.; Nørskov, J. K. Nat. Mater. 2006, 5, 909−913. (29) Nørskov, J. K.; Bligaard, T.; Logadottir, A.; Kitchin, J.; Chen, J. G.; Pandelov, S.; Stimming, U. J. Electrochem. Soc. 2005, 152, J23−J26. (30) Ryu, J.; Jung, N.; Jang, J. H.; Kim, H.-J.; Yoo, S. J. ACS Catal. 2015, 5, 4066−4074. (31) Huang, H.; Yu, C.; Yang, J.; Zhao, C.; Han, X.; Liu, Z.; Qiu, J. ChemElectroChem 2016, 3, 719−725. (32) Liang, H.; Meng, F.; Cabán-Acevedo, M.; Li, L.; Forticaux, A.; Xiu, L.; Wang, Z.; Jin, S. Nano Lett. 2015, 15, 1421−1427. (33) Sivanantham, A.; Ganesan, P.; Shanmugam, S. Adv. Funct. Mater. 2016, 26, 4661−4672. (34) Liang, H.; Li, L.; Meng, F.; Dang, L.; Zhuo, J.; Forticaux, A.; Wang, Z.; Jin, S. Chem. Mater. 2015, 27, 5702−5711. (35) Biesinger, M. C.; Payne, B. P.; Grosvenor, A. P.; Lau, L. W.; Gerson, A. R.; Smart, R. S. C. Appl. Surf. Sci. 2011, 257, 2717−2730.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.6b03803. Experimental details, additional supporting data, and comparison of catalytic performance (PDF)

■

Letter

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Husam N. Alshareef: 0000-0001-5029-2142 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The research reported in this publication was supported by funding from King Abdullah University of Science and Technology (KAUST). G

DOI: 10.1021/acs.nanolett.6b03803 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters (36) Grosvenor, A. P.; Biesinger, M. C.; Smart, R. S. C.; McIntyre, N. S. Surf. Sci. 2006, 600, 1771−1779. (37) Valeri, S.; Borghi, A.; Gazzadi, G.; Di Bona, A. Surf. Sci. 1999, 423, 346−356. (38) You, B.; Jiang, N.; Sheng, M.; Bhushan, M. W.; Sun, Y. ACS Catal. 2016, 6, 714−721. (39) Li, H.; Li, H.; Dai, W.-L.; Wang, W.; Fang, Z.; Deng, J.-F. Appl. Surf. Sci. 1999, 152, 25−34. (40) Conway, B.; Tilak, B. Electrochim. Acta 2002, 47, 3571−3594. (41) Wang, X.; Li, W.; Xiong, D.; Petrovykh, D. Y.; Liu, L. Adv. Funct. Mater. 2016, 26, 4067−4077. (42) Lukowski, M. A.; Daniel, A. S.; Meng, F.; Forticaux, A.; Li, L.; Jin, S. J. Am. Chem. Soc. 2013, 135, 10274−10277. (43) Lukowski, M. A.; Daniel, A. S.; English, C. R.; Meng, F.; Forticaux, A.; Hamers, R. J.; Jin, S. Energy Environ. Sci. 2014, 7, 2608− 2613. (44) Surendranath, Y.; Kanan, M. W.; Nocera, D. G. J. Am. Chem. Soc. 2010, 132, 16501−16509. (45) Xia, C.; Jiang, Q.; Zhao, C.; Hedhili, M. N.; Alshareef, H. N. Adv. Mater. 2016, 28, 77−85.

H

DOI: 10.1021/acs.nanolett.6b03803 Nano Lett. XXXX, XXX, XXX−XXX