- Email: [email protected]
A highly stable bifunctional catalyst based on 3D Co(OH)2@[email protected] towards overall water-splitting
Author’s Accepted Manuscript A highly stable bifunctional catalyst based on 3D Co(OH)2@NCNTs@NF towards overall watersplitting Pan Guo, Jian Wu, Xi-Bo Li, Jun Luo, Woon-Ming Lau, Hao Liu, Xue-Liang Sun, Li-Min Liu www.elsevier.com/locate/nanoenergy
PII: DOI: Reference:
S2211-2855(18)30101-0 https://doi.org/10.1016/j.nanoen.2018.02.032 NANOEN2519
To appear in: Nano Energy Received date: 8 January 2018 Revised date: 14 February 2018 Accepted date: 15 February 2018 Cite this article as: Pan Guo, Jian Wu, Xi-Bo Li, Jun Luo, Woon-Ming Lau, Hao Liu, Xue-Liang Sun and Li-Min Liu, A highly stable bifunctional catalyst based on 3D Co(OH)2@NCNTs@NF towards overall water-splitting, Nano Energy, https://doi.org/10.1016/j.nanoen.2018.02.032 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
A highly stable bifunctional catalyst based on 3D Co(OH)2@NCNTs@NF towards overall water-splitting Pan Guo‡1,2, Jian Wu‡4, Xi-Bo Li‡1, Jun Luo5, Woon-Ming Lau4, Hao Liu2*, Xue-Liang Sun6*, Li-Min Liu1,3* 1
Beijing Computational Science Research Center, Beijing 100193, China Chengdu Green Energy and Green Manufacturing Technology R&D Center, Chengdu Development Center of Science and Technology of CAEP, Chengdu, Sichuan, 610207, China 3 School of physics, Beihang University, Beijing 100191, P. R. China 4 Center for green Innovation, School of Mathematics and Physics, University of Science and Technology Beijing, Beijing 100083, China 5 Tianjin Key Laboratory of Advanced Functional Porous Materials and Center for Electron Microscopy, School of Materials Science and Engineering, Tianjin University of Technology, Tianjin 300384, China 6 Department of Mechanical and Materials Engineering, the University of Western Ontario, London, Ontario, N6A 5B9, Canada ‡ contributed equally to this work 2
Email: [email protected] or [email protected]; [email protected]; [email protected] Abstract Electrocatalysts with low overpotential and high stability are highly demanded in water-splitting system. The efficiency of water-splitting is largely restricted by the oxygen evolution reaction (OER). Here, we developed a two-step method to prepare 3D porous material through chemical vapour deposition and electrodeposition combined with the first-principles calculations. Ultrathin α-Co(OH)2 nanosheets grown on the combined substrate of N-doped carbon nanotubes (NCNTs) and nickel foam were fabricated to investigate their electrochemical behaviour. Because of the characteristics of the ultrathin, microporous α-Co(OH)2 and its derivatives, the 3D Co(OH)2@NCNTs@NF exhibits outstanding performance as a bifunctional catalyst for water-splitting. The overpotentials to achieve 10 mA cm-2 current density in 1 M KOH for OER and hydrogen evolution reaction (HER) are 270 mV and 170 mV, respectively. The as-prepared material exhibits superior stability, which generate 10 mA cm-2 current density in overall water-splitting over 600 hours without obvious degradation in 1 M KOH at voltage of 1.72 V vs. RHE. The first-principles calculations reveal that the N-doping not only can effectively enhance the interaction
between the substrate and active material (CoOOH), but also modulate the electronic structure of CoOOH to speed up the O2 releasing during the OER. Graphical abstract:
Keywords: Water-splitting; OER; HER; Electrodeposition; N-CNT
Introduction Searching for renewable energy productions is a rather critical issue to reduce the dependence on the fossil fuels[1-3]. Electrocatalytic water-splitting is one of the most promising ways to produce clean energy. In order to meet the requirements of the applications, it is long desired to develop cheap, non-toxic and efficient catalysts to stimulate the slow kinetics of water-splitting[4, 5]. The water-splitting process contains both OER and HER[6-8], and OER is much harsher than HER due to the complex four electrons process, which greatly restrains the efficiency and kinetics of the whole reaction[9, 10]. On the other side, some highly efficient OER catalysts do not exhibit good performance in HER because of the different
working medium. Therefore, developing highly-efficient, non-precious bifunctional catalysts which can be used in the same electrolyte is the ultimate way to meet industry demand. Cobalt is one of the earth-abundant elements and its corresponding hydroxide material is very easy to synthesize. Cobalt hydroxides are known to exist in two polymorphs, namely α- and β-Co(OH)2. Among them, the α-Co(OH)2 exhibits higher activity due to the fact that the layer-layer distance of α-Co(OH)2 is larger than that of β-Co(OH)2. However, most of the Co(OH)2 prepared are β-Co(OH)2[11, 12], so it is necessary to investigate the performance of α-Co(OH)2 as an electrocatalyst to catalyse water-splitting. Firstly, the NCNTs was deposited on the nickel foam (NF), then the Co(OH)2 nanosheets were grown on the NCNTs to form a binder-free 3D porous catalyst. This unique design has the following advantages: enhancing the accessible effective surface areas of the catalyst, facilitating escape of gases, improving the electron transfer and effectively avoiding aggregation/agglomeration during OER process[13, 14]. To the best of our knowledge, this work is the first systematic report focuses on α-Co(OH)2 to investigate its catalytic performance in water-splitting. Based on the aforementioned facts, a novel 3D self-supported porous material, namely Co(OH)2@NCNTs@NF, was fabricated, which exhibits a small overpotential of 270 mV in OER at 10 mA cm-2 current density in 1 M KOH over 200 h without any degradation. Meanwhile excellent catalytic performance towards HER reaction was obtained, which needs 170 mV overpotential to generate 10 mA cm-2 current density in 1 M KOH. We also constructed a two electrodes system to investigate catalytic performance in overall water-splitting and our catalyst generate 10 mA cm-2 current density in 1 M KOH with only 490 mV overpotential needed and exhibit such an amazing stability in 10 mA cm-2 water-splitting process (>600 h).
Experiment results SEM and TEM images show that the NCNTs were uniformly distributed over the entire nickel foam (Fig.
S1a). The nanotubes intertwined with each other and connected to the nickel foam directly, forming a binder-free 3D porous skeleton (Fig. S1b). From the magnified image (Fig. S1c and S1d), the NCNTs exhibits a bamboo-like tubular structure and a rough surface, which are the typical characteristics of NCNTs[15-17]. This 3D structure can facilitate fast transportation of electrons between NF and NCNTs as well as effective mass exchange in electrolyte, and rough surface of NCNTs are beneficial for electrodepositing. Following this, the Co(OH)2 nanosheets were evenly deposited onto the surface of each nanotubes, forming a blue thin film on the surface of black NCNTs (Fig. S2). The SEM images (Fig. 1a and S3) show that the Co(OH)2 nanosheets interconnected with each other to form a nanosheet array structure. The low magnification SEM image presented in Fig. 1b shows that the Co(OH)2 nanosheets completely cover over the NCNTs. As a comparison, Co(OH)2 nanosheets were also electrodeposited onto bare NF under the same experimental conditions. As shown in Fig. 1c, the nanosheets on NCNTs and NF are greatly different. The thickness and lateral size of the nanosheets on NF are obviously thicker and smaller than those on NCNTs, respectively, which clearly verities the importance of NCNTs for loading materials[18-20]. The TEM image (Fig. 1d) further verifies the intimate and direct contact between Co(OH)2 and NCNTs. The TEM image (Fig. S4) of the marked area in Fig. 1d exhibits that the thickness of the nanosheet is around 6 nm. As displayed in TEM image of Fig. 1e, the nanosheets are nearly transparent, indicating their ultrathin feature. The spacings of (002) and (100) atomic planes are determined to be 0.46 nm and 0.27 nm (Fig. 1f), respectively, which is consistent with the data of α-Co(OH)2[21]. Meanwhile, the XRD result further suggests that the electrodeposited material is α-Co(OH)2 (see in Fig. S5a). The selected-area electron diffraction pattern shown in inset of Fig. 1f exhibits a diffused spot pattern, further confirming the polycrystalline nature of the electrodeposited Co(OH)2. The elemental mapping of Co(OH)2@NCNT (Fig. 1g) shows a uniform distribution of C, Co, O and N elements in the composite. Among them, C and N elements exist in the region of NCNT, while Co and O elements distributed over
the entire region, which again verifies the designed structure.
Fig. 1 | The morphology characterization of the as-prepared material (a) SEM image of Co(OH)2 nanosheets on NCNTs. (b) low magnification SEM image of Co(OH)2@NCNTs. (c) The SEM image of Co(OH)2 electrodeposited on NF. (d) TEM image of Co(OH)2@NCNTs. (e) TEM image of Co(OH)2 nanosheets. (f) HRTEM image of Co(OH)2 nanosheets (inset: SAED pattern of the nanosheets). (g) Elemental mapping of C, N, Co and N in Co(OH)2@NCNTs composite.
In order to obtain a more stereoscopic and realistic image of the Co(OH)2@NCNTs, 3D reconstruction method were taken to restore the original 3D structure of the as-prepared material. The original image (Fig. 2a) exhibits clear stereoscopic structure of Co(OH)2 nanosheets. The nanosheets grow from NCNT axis and have a rough surface which exposes more active area than smooth surface. Meanwhile, the nanosheets are not stacked to each other, thus the open spaces between nanosheets ensure enough contact between electrolyte and Co(OH)2 surface. Fig. 2b displays the transparent mode of the original 3D structure, in which the CNT, doped-N and Co(OH)2 nanosheets are clearly seen. The
N-doped CNTs possess concave-convex structure (seen in Fig. S6) which is beneficial for robust adhesion between Co(OH)2 nanosheets and NCNTs. The top view image (shown in Fig. 2c) of the truncated section in Fig. 2b implies that the nanosheets grow radially from the NCNT axis and the emerged areas between nanosheets and N may be the original nucleation sites.
Fig. 2 | The 3D reconstruction characterization of the Co(OH)2 nanosheets on N doped CNT.
The chemical composition of Co(OH)2@NCNT@NF is further determined by X-ray photoelectron spectroscopy (XPS). The results are shown in Fig. 3a and b. There are two peaks at 781.4 and 797.3 eV accompanied by two satellite peaks at 786.0 and 803.0 eV in the Co 2p spectrum (Fig. 3a), indicating the oxidation state of Co is +2. The XPS spectrum of O 1s (Fig. 3b) exhibits single peak at 531.2 eV, corresponding to the bound hydroxide groups[22]. These results collectively confirm the synthesis of Co(OH)2. The XPS spectrum of N 1s (Fig. S7) can be deconvoluted into pyridine N (398.6 eV), oxidized N (402.3 eV) and graphite N (401.0 eV), among which pyridinc and graphitic N are considered as the OER active sites[23].
Fig. 3 | The XPS spectra of prepared material (a) Co 2p and (b) O 1s XPS spectra of Co(OH)2@NCNTs@NF before being tested. (c) Co 2p and (d) O 1s XPS spectra of Co(OH)2@NCNTs@NF after stability test.
Electrochemical Results A series of electrochemical experiments were carried out in 1 M KOH to investigate the performance of the as-prepared catalysts. We made three batches of Co(OH)2@NCNTs@NF with different electrodeposition time to determine the suitable time. Taking the onset potential and overpotential at 10 mA cm-2 as judging criterion, the 20 minutes sample shows the best performance in both onset potential and overpotential (Fig. S8), so in the following experiment 20 minutes was taken as the electrodeposition time. At first, the linear sweep voltammetry data of Co(OH)2@NCNTs@NF, Co(OH)2@NF, NCNTs@NF and NF were recorded. As shown in Fig. 4a, the substrate NF shows a negligible current in the voltage range from 1.40 V to 1.60 V. Thus, it is a suitable substrate for studying the performance of high active materials in low overpotential ranges. Co(OH)2@NF and NCNTs@NF also exhibit relatively unsatisfactory catalytic activity in this voltage range. The as-prepared catalyst, Co(OH)2@NCNTs@NF, manifests superior advantages not only in onset potential but also in OER current at any given voltages. Besides, the current density of
Co(OH)2@NCNTs@NF at 1.50 V is higher than the sum of both NCNTs@NF and Co(OH)2@NF at the same voltage. Therefore, a synergistic effect must exist between Co(OH)2 and NCNTs to promote the catalytic activity of Co(OH)2@NCNTs@NF. The LSV data of Co(OH)2@NCNTs@NF exhibits two peaks at 1.10 V and 1.43 V[12, 24, 25], which corresponds to the transformation of Co2+/Co3+ and Co3+/Co4+, respectively. The formation of Co4+ is critical for improving the catalytic activity of Cobalt-based materials because the Co4+ is considered as the main active sites for OER[26-28]. As shown in the LSV data of Co(OH)2@NF, one peak exists at 1.13 V, while the peak corresponding to transformation of Co3+/Co4+ is negligible, which is due to the competition reactions[29]. The negligible peak of transformation of Co3+/Co4+ indicates that little Co4+ formed in Co(OH)2@NF comparing to Co(OH)2@NCNTs@NF, consequently, the performance of Co(OH)2@NF is relatively unsatisfactory. As shown in the inset of Fig. 4a, the capacitance reaction is completed before the OER, therefore the current after 1.47 V mainly comes from OER. The as-prepared Co(OH)2@NCNTs@NF only needs 270 mV overpotential to generate 10 mA cm-2 current density, which is among the limited materials with overpotential lower than 300 mV to generate 10 mA cm-2 current density[14,
30-34]
. Such a small
overpotential is the lowest one according to our knowledge in either Co or Co(OH) 2-based materials. The performance in high current density conditions is also excellent. As shown in Fig. 4a, it can reach 100 mA cm-2 current density at the overpotential of 410 mV, which is superior in current cobalt-based catalysts. As shown in Fig. 4b, the Tafel curves of the three materials are plotted[30], and the Tafel slope of Co(OH)2@NCNTs@NF is clearly lower than those of Co(OH)2@NF and NCNTs@NF. Considering that lower Tafel slope means faster catalytic kinetics[31, 35, 36]; thus, Co(OH)2@NCNTs@NF exhibits a better catalytic performance than their counterparts. Although the Tafel slope of Co(OH)2@NCNTs@NF is not comparable to the state-of-the-art precious metal catalysts, the relatively low value of 72 mV/dec also shows advantage in transition metal-based catalysts. Long time duration of the catalysts in the working life span is another important parameter. The
overall water-splitting is usually carried out at a constant current or voltage. In order to mimic a more realistic working condition, the chronopotentiometry was used to evaluate the catalyst stability. The chronopotentiometry test was taken at 10 mA cm-2, Co(OH)2@NCNTs@NF exhibits an amazing stability over 200 hours without any degradation (see Fig. 4c). In order to mimic a harsher condition, another stability test was taken. At this test, the sample was firstly tested for 100 hours at 10 mA cm-2, then for another 100 hours test at 20 mA cm-2. The Co(OH)2@NCNTs@NF maintains a nearly constant voltage for the final 100 hours at 20 mA cm-2 (Fig. 4d). Such a long and steady stability performance excludes the corrosion of carbon nanotubes and also verifies the strong structure of our catalyst[33]. To take a more detailed comparison between the catalyst before and after testing, the LSV data of the catalyst after testing is recorded. The comparison data were shown in Fig. 4e, the LSV curve of the catalyst after testing at 10 mA cm-2 for 100 hours is similar to that after another 100 hours at 20 mA cm-2 except a little degradation for the latter one at voltage > 1.60 V. The efficient HER catalysts usually work at acid medium, which greatly limits the application of transition metal oxides and hydroxides, and thus, it is rather challenging to combine OER and HER catalysts in one electrolyte medium[32, 34]. Therefore, efficient bifunctional catalysts working in one medium are extremely desired. The HER performance of Co(OH)2@NCNTs@NF is also examined in 1 M KOH. The optimal electrodeposition time for catalyst used in HER is determined to be 15 minutes (see Fig. S9). The LSV curve presented in Fig. 4f clearly demonstrates the advantages of Co(OH)2@NCNTs@NF over their counterparts. The Co(OH)2@NCNTs@NF electrode reaches 10 mA cm-2 current density at a small overpotential of 170 mV. The chronopotentiometry measurement was also taken to investigate its stability. As shown in Fig. S10, the Co(OH)2@NCNTs@NF electrode undergoes an activation process during stability test. The overpotential needed to generate 10 mA cm-2 current density gradually decreased to 115 mV after 120 h testing and was stable around this potential during the following 400 h testing.
Owing to the excellent catalytic performance, the Co(OH)2@NCNTs@NF can be used as OER and HER electrodes to catalyse overall water-splitting. As shown in Fig. S11, Co(OH)2@NCNTs@NF exhibits an excellent performance with a voltage of 1.72 V needed to generate 10 mA cm -2 current density in overall water-splitting. Chronopotentiometry test was further carried out to measure the durability of the as-prepared catalyst. As presented in Fig. 4g, the Co(OH)2@NCNTs@NF electrode exhibits the good stability over the whole testing period (600 h), clearly indicating its attractiveness in practical applications.
Fig. 4 | Electrochemical characterizations. (a) Linear sweep voltammetry curves of the Co(OH)2@NCNTs@NF electrode and their counterparts for OER in 1 M KOH solution at 2 mV s-1. (b) Tafel plots of the three electrodes in 1 M KOH solution at 2 mV s-1. (c) Chronopotentiometry curves obtained with the Co(OH)2@NCNTs@NF act as OER electrode in 1 M KOH at constant current density of 10 mA cm-2. (d) Chronopotentiometry curves taken at 20 mA cm-2 in 1 M KOH after the Co(OH)2@NCNTs@NF electrode being tested in 1 M KOH for 100 h at 10 mA cm-2. (e) Linear sweep voltammetry curves of the Co(OH)2@NCNTs@NF electrode after Chronopotentiometry test. (f) Linear sweep voltammetry curves of
the Co(OH)2@NCNTs@NF electrode and their counterparts for HER in 1 M KOH solution at 2 mV s-1. (g) Chronopotentiometry curves taken at 10 mA cm-2 in 1 M KOH toward overall water-splitting.
XRD and XPS were carried out on Co(OH)2 after cycling, and interestingly, the results show that the α-Co(OH)2 transformed into ϒ-CoOOH after cycling ( Fig. 3c and 3d). As shown in Fig. S5b, the XRD diffraction pattern belongs to ϒ-CoOOH[37]. Although there is no specific PDF card in JCPDS corresponding to ϒ-CoOOH, the peaks at about 13.66°, 26.98°, 37.88° and 39.73° approximately match the (003), (006), (101) and (105) peaks of ϒ-NiOOH[38, 39]. As shown in Fig. 3c, the XPS spectrum of Co after cycling is different from that of the pristine one. The best deconvolution of Co 2p can be accomplished by the assumption that Co2+ and Co3+ coexist in the material. As shown in Fig. 3c, the dominant peaks at 780.0 and 795.0 eV belong to the Co3+ 2p3/2 and 2p1/2, respectively. The two satellite peaks at 790.0 and 805.0 eV further indicates the formation of CoOOH during OER process. Also, the XPS spectrum of O 1s (Fig. 3d) is different from that of the original one with two peaks located at 530.9 and 529.1 eV, which are attributed to the O from the OH and oxide ions[40]. The XPS spectrum of O 1s in CoOOH is different from that of either Co(OH)2 or Co3O4, and these latter two materials have the main peak located in 531.2 and 530.0 eV, respectively[40]. The difference of O 1s spectrum in CoOOH with those in Co(OH)2 and Co3O4 clearly confirms the formation of CoOOH.
Discussion First of all, the ultrathin Co(OH)2 nanosheets contain many micropores (Fig. 5a) which can expose more grain boundaries to participate in OER. As discussed above, the α-Co(OH)2 is different from the widely explored β-Co(OH)2, whose interlayer distance is about twice larger. The relatively large interlayer distance provides more active sites and the thin nanosheets shorten the diffusion length of electrolyte. Meanwhile, this property is inherited by its derivative, ϒ-CoOOH, which also exhibits the relatively large
interlayer distance. Besides that, the surface CoOOH formed during OER process on the Co(OH) 2 nanosheets is a better conductor for electrons[41, 42]. Secondly, the unique spatial structure plays a key role in catalysing OER. As shown in SEM image in Fig. S1b, the as-prepared catalyst featured a woven-like intertwined 3D network structure, which are beneficial for fast transportation of electrolyte and electrons. What’s more, the tubular structure and rough surface, typical characters of NCNTs, can provide more nucleate sites for uniformly electrodepositing Co(OH)2. The interaction between NCNTs and Co(OH)2 is quite strong and no visible voids exist between NCNTs and Co(OH)2 (Fig. 1d). This intimate contact is very important to retain a robust structure and strong synergistic effect between N and Co. It is worth mentioning that it takes 20 minutes ultrasonication to disperse the catalyst during TEM sample preparation, which again confirms the robust structure of the catalyst. The structure after the stability test were also examined. As shown in Fig. 5b and 5c, SEM and TEM images indicate that the catalyst maintains its original structure after 200 hours testing.
Fig. 5 | The morphology characterization of the Co(OH)2@NCNTs. (a) Magnified TEM image of Co(OH)2 nanosheets. (b) SEM and (c) TEM images of the Co(OH)2@NCNTs@NF electrode after stability test at 10 mA cm-2 for 200 h.
Last, the role of N dopant is also very important. As analysed in XPS section, there are three kinds of N, namely pyridine N, oxidized N and graphite N, among which pyridinic and graphitic N atoms are considered to be the active sites for OER reaction[23]. To understand why the N-doping can greatly improve the performance of the CoOOH in OER, first-principles calculations were carried out. First of all, the most stable configurations of monolayer β-CoOOH and the heterostructure between monolayer
β-CoOOH and graphene with one pyridinic N defect are examined to mimic the interaction between the CoOOH and NCNT. The relative stability of several different monolayer CoOOH configurations are calculated. Four typical monolayer CoOOH configurations are shown in Fig. S13. The most stable configuration of monolayer CoOOH (denoted as CoOOH@) is shown in Fig. S13(a), and the main difference from others is that H atoms are distributed equally on the oxygen atoms. Based on the most stable configuration of monolayer CoOOH, the heterostructure between monolayer CoOOH and graphene with the pyridinic N defect are constructed. The corresponding configurations and the relative energies are shown in Fig. S13(e) – (f). The heterostructure in Fig. S13(e) is the most stable one, in which the N atom bonds with the cobalt atoms in CoOOH (denoted as CoOOH@N1). Based on the most stable configuration of CoOOH@ and CoOOH@N, the OER processes were calculated. In order to know the effect of N concentration on the OER, the system containing two N atoms was also considered (denoted as CoOOH@N2, Fig. S13(h)). Our calculation results show that the formation energies of oxygen vacancy in CoOOH@, CoOOH@N and CoOOH@N2 are 2.56, 2.19 and 0.19 eV per oxygen vacancy, respectively. This trend indicates that the oxygen vacancy is relatively easier to form in heterostructure with one N defect than the one in the pure monolayer CoOOH, and even easier in heterostructure with more N defects. In the following, the OER activities of CoOOH@, CoOOH@N1 and CoOOH@N2 are calculated. Fig. 6 shows the free energy changes of OER steps on CoOOH@, CoOOH@N1 and CoOOH@N2, and the corresponding intermediate species of OER are shown in Fig. S14 and Fig. S15. The first principles calculations indicate that the last step of O2 formation (*OOH + → *O2(g) + H+ + e-) is the potential determining step (PDS) on CoOOH@ and CoOOH@N1, with the free energy barriers of 2.83 eV and 0.56 eV at U = 1.23 eV vs SHE, respectively. The reduced overpotential for heterostructure exhibits that CoOOH@N1 show better OER performance than the pure CoOOH@. Meanwhile, with the increase of the concentration of N, the overpotential of OER in CoOOH@N2 become 0 eV. These results clearly
suggest that the N dopant can greatly enhance OER activity.
Fig. 6 | Free energy changes diagram at pH = 13, T = 298 K for the four steps of OER at U= 0 eV vs SHE (black solid line) at equilibrium potential U = 1.23 eV vs SHE (red dashed line) on (a) CoOOH@, monolayer CoOOH; (b) CoOOH@N1, hetrostrucuture of CoOOH binding with one pyridinic N defect; (c) CoOOH@N2: hetrostrucuture of CoOOH binding with two pyridinic N defects. The overpotential are in parentheses and the negative value in the (c) indicates the overpotential is 0 eV.
In order to unveil the origin of the high OER activity of the heterostructure, the electron transfers from the monolayer CoOOH to the graphene were calculated. For the CoOOH@, there is no electron transfer considering the system does not contain the interface. While, the corresponding values are 0.16 e for CoOOH@N1 and 0.58 e for CoOOH@N2, respectively. In addition, the charge density difference of CoOOH@N1 is also given to show the distribution of the electron. As show in Fig. S16, it clearly shows that there is electron transfer from Co atoms in CoOOH_Ov to N atom in pyridinic N defect in graphene. As shown in Fig. 6, with increase of the electron transfer, the free energy of adsorption of *OOH on CoOOH, △G(OOH), becomes weaker: 0.09 eV for CoOOH@, 2.36 eV for CoOOH@N1, 3.69 eV for CoOOH@N2, which facilitates the O2 formation step (*OOH + → *O2(g) + H+ + e-). Meanwhile, the adsorption ability of *O is correlated with the formation energy of oxygen vacancy for CoOOH@, CoOOH@N, and CoOOH@N2. The smaller the formation energy of oxygen vacancy in CoOOH, the smaller △G(O) for the corresponding OER. Further, the OER activities of CoOOH@, CoOOH@N, and CoOOH@N2 are greatly related with the
oxidation state of the Co atoms around the oxygen vacancy. The total charge state of three Co atoms nearby the oxygen vacancy in CoOOH@, CoOOH@N, CoOOH@N2 are +1.76 e, +2.25 e, and +2.41 e, respectively. As shown in Fig. 7, CoOOH@N and CoOOH@N2 loses more electrons to pyridinic N defect than CoOOH@, the Co atoms in CoOOH@N and CoOOH@N2 behave higher charge state, which greatly decrease the ability of strongly adsorbing O or OOH by transferring the extra electron to O or OOH. In all, the electrons transfer in the heterostructure makes the cobalt atoms around oxygen vacancy behave as the high oxide state, effectively decreasing the adsorption ability of *OOH and the overpotential of OER.
Fig. 7 | The relationship between electron transfer in the heterostructure and △G(O) (blue square), △G(OOH) (red round) and Gf(Ov) (black triangle) for CoOOH@, CoOOH@N, and CoOOH@N2.
Conclusions Owing to the merits of nickel foam and N-doped carbon nanotubes, we take a combined method including CVD and electrodeposition to prepare 3D porous catalyst Co(OH)2@NCNTs@NF, which is confirmed to be an excellent catalyst with small overpotential and long stability. This material can generate OER current density of 10 mA cm-2 at voltage of 1.50 V vs. RHE, corresponding to an
overpotential of 270 mV, which is among the limited materials with an overpotential lower than 300 mV to reach 10 mA cm-2 current density. The stability of our catalyst is also significantly improved with a working life span lasting over 200 hours without degradation. This excellent material also catalyse HER with good performance. It only needs 170 mV overpotential to generate 10 mA cm -2 current density in 1 M KOH. Benefitting from the excellent performance towards OER and HER, our material catalyse overall water-splitting at 1.72 V vs. RHE to generate 10 mA cm-2 current density and this performance can maintain over 600 hours without degradation. Therefore our preparation method can be expected to be adopted in industry application.
Experimental section Preparation of N-doped CNTs on nickel foam We use a chemical vapor deposition (CVD) method developed in our lab to prepare N-doped CNTs on nickel foam[43-45]. The original nickel foam will become brittle after annealing, which is unfavorable for acting as a substrate material. After a series of trial and error, we found that the pressed nickel foam can sustain its ductility after annealing. Thus, all the nickel foam were pressed before using. Before the deposition process, the nickel foam were cut into 1 cm × 1.5 cm squares and pressed to 0.16 mm thickness, and then put into a 6 M HCl solution to sonicate for 30 minutes. After that the nickel foam were rinsed with copious deionized water (DI) and ethanol to remove the residue impurity. The obtained nickel foam were put in an electronic vacuum oven for 12 hours to vapor the trace amount of liquid. Following that, the nickel foam were used as the substrate to load N-doped CNTs. Briefly, the nickel foam were placed in the middle of a quartz tube. The system was purged with high purity argon gas with a flow rate of 800 sccm for 30 minutes to ensure an inert environment. After that the furnace was heated in a programmable step to 900 C in 30 minutes and then kept at this temperature for another 15 minutes. Ferrocene, ethylene gas and melamine which acts as the catalyst, carbon source and nitrogen
source, respectively, was introduced into the tube lasting for 15 minutes to deposition N-doped CNTs. Finally after turning down the ethylene gas, the furnace was left to cool down to room temperature, during which the argon gas was continue purged to protect the obtained material. The obtained material was donated as NCNTs@NF. Preparation of Co(OH)2@NCNTs@NF The electrodeposition process was completed in a standard three electrode system with the NCNTs@NF as the working electrode, platinum mesh as the counter electrode and saturated calomel electrode (SCE) as the reference electrode. The Co(OH)2 nanosheets were electrodeposited onto the NCNTs@NF substrate in an aqueous solution containing 6 mmol Co(NO3)2•6H2O at a constant cathodic current density of 1.5 mA cm-2 for 20 minutes. Following that, the obtained Co(OH)2@NCNTs@NF was washed with copious water and ethanol for several times repeatedly. Finally, the as-prepared sample was vacuum dried at 60 C for 2 hours. The loading mass of Co(OH)2 nanosheets was determined by the mass difference between the substrate before and after electrodeposition using a high precise electronic balance. The average mass was around 0.72 mg cm-2. We also electrodeposited Co(OH)2 onto bare nickel foam in the same condition except for the substrate as a comparison to study the possible synergistic effect between Co(OH)2 and N-doped CNT.
Material characterization The phase purity and crystallographic information were obtained by using the X-ray diffraction (XRD, D/max 2200/PC, Rigaku, 40 kV, 20 mA, Cu K α radiation, λ = 1.5406 Å). The surface morphology and elemental composition of the sample were investigated by scanning electron microscopy (SEM, Hitachi S-5200). The more detailed microstructure was characterized by using high resolution transmission electron microscope (HRTEM, FEI, Tecnai F20) equipped with energy dispersive X-ray spectroscopy elemental mapping. The 3D reconstruction were carried out by Transmission Electron Microscope with A
Probe Corrector (Titan Cubed Themis G2 300).
Electrochemical measurements All the involved electrochemical measurements were carried out in a CHI 760E electrochemical workstation in 1 M KOH aqueous solution at room temperature. The prepared Co(OH)2@NCNTs@NF, Co(OH)2@NF and NCNTs@NF were used as the working electrodes directly without any further treatment. A platinum mesh and saturated calomel electrode was used as the counter and reference electrode, respectively. The SCE was calibrated using a method described in Dai’s work[46] and the relationship between RHE and SCE in 1 M KOH is determined to be: E RHE = ESCE + 1.062 V (see in Fig. S12). Before any measurements the three electrode system was purged with high purity O 2 gas for 20 minutes and O2 was maintained during the whole testing to ensure the equilibrium of O2/H2O at 1.23 vs. RHE[30]. The electrochemical data were recorded only after the system was cycled to reach a stable cyclic voltammogram. The cyclic voltammogram (CV) and linear sweep voltammetry (LSV) was performed at a scan rate of 50 mV s-1 and 2 mV s-1, respectively. The scan rate choice of LSV is slow enough to minimize the capacitance current and obtain accurate polarized OER current curve. All the obtained results are not compensated by iR drop because the internal resistance is very small. Chronopotentiometry measurement was carried out in the same experimental setup with no iR drop compensation.
First-principles calculations. All calculations were performed using DFT methods, as implemented in the CP2K/Quickstep package[47]. The exchange correlation potential was described by the generalized-gradient approximation (GGA) with the spin-polarized functional of Perdew-Burke-Ernzerh (PBE)[48]. A cutoff energy of 500 Ry was chose for the wave functions. Core electrons were modeled by norm-conserving pseudopotentials[49] with 17, 6, 1, 4 and 5 valence electrons of Co, O, H, C and N, respectively. Only the Γ-point was used for Brillouin zone integration. The DFT+U method, based on the
Mullikan 4d state population analysis, was used to describe the Co 4d electrons. A U value of 3.52 eV for Co was used in all calculations[50]. Besides, van der Walls interaction was also considered with the D2 [51] in our DFT calculations. The details of the models used are as discussed as follows. In the monolayer CoOOH models, the (2×3) supercell were explored for further calculations. In the heterostructure model, the (2×3) monolayer CoOOH and the (4√2×5√2)R30。graphene with pyridinic N defects are perpendicular to each other. In order to explore effect of the density of pridinic defect on OER, the heterostructure models with one and two pridinic defects in graphene of the heterstructure were established. In additional, there was one oxygen vacancy in both in monolayer CoOOH model and heterostructure models.
Free Energy calculations. As proposed in the previous work, it is more convenient to model the thermochemistry of OER at acid condition[50, 52], where the steps involving electron transfer in step (1)─(4) are modified as follow *+ H2O +→OH*+H++ e-
(1)
OH*→O *+ H++e-
(2)
O*+ H2O →OOH*+ H++e-
(3)
OOH*+→*+O2(g)+ H++e-
(4)
where * stands for the activity site on the catalyst’s surface. The change of Gibbs free energy in each reaction step could be expressed as: △G = △E + △ZPE + △H - T△S+△GH+(pH), where △E, △ZPE, △H, T△S , △GH+(pH)indicate the changes of electronic energy, zero point energy, the enthalpy, the entropy, free energy change of proton relative to pH = 0, respectively. And more pH effect on H+ is considered as △GH+(pH) = -kBT ln(10) × pH[53]. In our calculations, the zero point energy was not considered. Using the standard hydrogen electrode (T = 298.15 K, P = 1 bar, pH = 0), the Gibbs energy of the proton/electron could be expressed as that of H2, GH++ Ge-=1/2 GH2. As the difficulty of determining
the energy of O2 within GGA-DFT frame, the free energy of O2 could be induced by equation: GO2(g) = 2 GH2O - 2 GH2 + 4.92 eV.
References [1] M. Lefevre, E. Proietti, F. Jaouen, J.P. Dodelet, Science, 324 (2009) 71-74. [2] C. Liu, F. Li, L.P. Ma, H.M. Cheng, Adv Mater, 22 (2010) E28-62. [3] K.S. Joya, Y.F. Joya, K. Ocakoglu, R. van de Krol, Angew Chem Int Ed Engl, 52 (2013) 10426-10437. [4] A. Aijaz, J. Masa, C. Rösler, W. Xia, P. Weide, A.J.R. Botz, R.A. Fischer, W. Schuhmann, M. Muhler, Angewandte Chemie International Edition, 55 (2016) 4087-4091. [5] N.S. Lewis, D.G. Nocera, Proceedings of the National Academy of Sciences, 103 (2006) 15729-15735. [6] M.G. Walter, E.L. Warren, J.R. McKone, S.W. Boettcher, Q. Mi, E.A. Santori, N.S. Lewis, Chem Rev, 110 (2010) 6446-6473. [7] C.C. McCrory, S. Jung, I.M. Ferrer, S.M. Chatman, J.C. Peters, T.F. Jaramillo, J Am Chem Soc, 137 (2015) 4347-4357. [8] F. Song, X. Hu, Nat Commun, 5 (2014) 4477. [9] R.D.L. Smith, M.S. Prévot, R.D. Fagan, Z. Zhang, P.A. Sedach, M.K.J. Siu, S. Trudel, C.P. Berlinguette, Science, 340 (2013) 60-63. [10] T.R. Cook, D.K. Dogutan, S.Y. Reece, Y. Surendranath, T.S. Teets, D.G. Nocera, Chem Rev, 110 (2010) 6474-6502. [11] J.-X. Feng, L.-X. Ding, S.-H. Ye, X.-J. He, H. Xu, Y.-X. Tong, G.-R. Li, Advanced Materials, 27 (2015) 7051-7057. [12] J.A. Koza, C.M. Hull, Y.-C. Liu, J.A. Switzer, Chemistry of Materials, 25 (2013) 1922-1926. [13] X. Yu, M. Zhang, J. Chen, Y. Li, G. Shi, Advanced Energy Materials, 6 (2016) 1501492-1501501. [14] X. Lu, C. Zhao, Nat Commun, 6 (2015) 6616.
[15] X. Xu, S. Jiang, Z. Hu, S. Liu, ACS Nano, 4 (2010) 4292-4298. [16] Z. Chen, D. Higgins, Z. Chen, Carbon, 48 (2010) 3057-3065. [17] D. Geng, H. Liu, Y. Chen, R. Li, X. Sun, S. Ye, S. Knights, Journal of Power Sources, 196 (2011) 1795-1801. [18] Y. Liang, Y. Li, H. Wang, H. Dai, J Am Chem Soc, 135 (2013) 2013-2036. [19] H. Zhu, J. Zhang, R. Yanzhang, M. Du, Q. Wang, G. Gao, J. Wu, G. Wu, M. Zhang, B. Liu, J. Yao, X. Zhang, Adv Mater, 27 (2015) 4752-4759. [20] M.-R. Gao, Y.-F. Xu, J. Jiang, Y.-R. Zheng, S.-H. Yu, J Am Chem Soc, 134 (2012) 2930-2933. [21] Z. Liu, R. Ma, M. Osada, K. Takada, T. Sasaki, J Am Chem Soc, 127 (2005) 13869-13874. [22] H.B. Li, M.H. Yu, X.H. Lu, P. Liu, Y. Liang, J. Xiao, Y.X. Tong, G.W. Yang, ACS Appl Mater Interfaces, 6 (2014) 745-749. [23] Y. Zhao, R. Nakamura, K. Kamiya, S. Nakanishi, K. Hashimoto, Nat Commun, 4 (2013) 2390. [24] M.A. Sayeed, T. Herd, A.P. O'Mullane, Journal of Materials Chemistry A, 4 (2016) 991-999. [25] Y.-C. Liu, J.A. Koza, J.A. Switzer, Electrochimica Acta, 140 (2014) 359-365. [26] B.E. Conway, T.C. Liu, Materials Chemistry and Physics, 22 (1989) 163-182. [27] G.W. Simmons, E. Kellerman, H. Leidheiser, Journal of The Electrochemical Society, 123 (1976) 1276-1284. [28] G.W. Simmons, A. Vértes, M.L. Varsányi, H. Leidheiser, Journal of The Electrochemical Society, 126 (1979) 187-189. [29] J.A. Koza, C.M. Hull, Y.-C. Liu, J.A. Switzer, Chemistry of Materials, 25 (2013) 1922-1926. [30] S. Chen, J. Duan, P. Bian, Y. Tang, R. Zheng, S.-Z. Qiao, Advanced Energy Materials, 5 (2015) 1500936-1500943. [31] X. Long, J. Li, S. Xiao, K. Yan, Z. Wang, H. Chen, S. Yang, Angew Chem Int Ed Engl, 53 (2014) 7584-7588.
[32] H. Wang, H.W. Lee, Y. Deng, Z. Lu, P.C. Hsu, Y. Liu, D. Lin, Y. Cui, Nat Commun, 6 (2015) 7261. [33] M. Gong, Y. Li, H. Wang, Y. Liang, J.Z. Wu, J. Zhou, J. Wang, T. Regier, F. Wei, H. Dai, J Am Chem Soc, 135 (2013) 8452-8455. [34] H. Jin, J. Wang, D. Su, Z. Wei, Z. Pang, Y. Wang, J Am Chem Soc, 137 (2015) 2688-2694. [35] X. Li, Y. Fang, X. Lin, M. Tian, X. An, Y. Fu, R. Li, J. Jin, J. Ma, Journal of Materials Chemistry A, 3 (2015) 17392-17402. [36] S. Liu, L. Li, H.S. Ahn, A. Manthiram, Journal of Materials Chemistry A, 3 (2015) 11615-11623. [37] Z. Chang, H. Li, H. Tang, X.Z. Yuan, H. Wang, International Journal of Hydrogen Energy, 34 (2009) 2435-2439. [38] X.-Z. Fu, Y.-J. Zhu, Q.-C. Xu, J. Li, J.-H. Pan, J.-Q. Xu, J.-D. Lin, D.-W. Liao, Solid State Ionics, 178 (2007) 987-993. [39] L. Jun, L. Rong, W. Jianming, S. Hang, Journal of Power Sources, 79 (1999) 86-90. [40] J. Yang, H. Liu, W.N. Martens, R.L. Frost, The Journal of Physical Chemistry C, 114 (2010) 111-119. [41] V. Pralong, A. Delahaye-Vidal, B. Beaudoin, J.B. Leriche, J.M. Tarascon, Journal of The Electrochemical Society, 147 (2000) 1306. [42] J. Huang, J. Chen, T. Yao, J. He, S. Jiang, Z. Sun, Q. Liu, W. Cheng, F. Hu, Y. Jiang, Z. Pan, S. Wei, Angew Chem Int Ed Engl, 54 (2015) 8722-8727. [43] H. Liu, Y. Zhang, R. Li, X. Sun, S. Désilets, H. Abou-Rachid, M. Jaidann, L.-S. Lussier, Carbon, 48 (2010) 1498-1507. [44] R. Mi, S. Li, X. Liu, L. Liu, Y. Li, J. Mei, Y. Chen, H. Liu, H. Wang, H. Yan, W.-M. Lau, Journal of Materials Chemistry A, 2 (2014) 18746-18753. [45] J. Wu, P. Guo, R. Mi, X. Liu, H. Zhang, J. Mei, H. Liu, W.-M. Lau, L.-M. Liu, Journal of Materials Chemistry A, 3 (2015) 15331-15338. [46] Y. Liang, Y. Li, H. Wang, J. Zhou, J. Wang, T. Regier, H. Dai, Nat Mater, 10 (2011) 780-786.
[47] J. VandeVondele, M. Krack, F. Mohamed, M. Parrinello, T. Chassaing, J. Hutter, Computer Physics Communications, 167 (2005) 103-128. [48] J.P. Perdew, A. Ruzsinszky, G.I. Csonka, O.A. Vydrov, G.E. Scuseria, L.A. Constantin, X. Zhou, K. Burke, Physical Review Letters, 100 (2008) 136406. [49] S. Goedecker, M. Teter, J. Hutter, Physical Review B, 54 (1996) 1703-1710. [50] M. Bajdich, M. García-Mota, A. Vojvodic, J.K. Nørskov, A.T. Bell, Journal of the American Chemical Society, 135 (2013) 13521-13530. [51] S. Grimme, J. Antony, S. Ehrlich, H. Krieg, The Journal of Chemical Physics, 132 (2010) 154104. [52] J.K. Nørskov, J. Rossmeisl, A. Logadottir, L. Lindqvist, J.R. Kitchin, T. Bligaard, H. Jónsson, The Journal of Physical Chemistry B, 108 (2004) 17886-17892. [53] M. García-Mota, M. Bajdich, V. Viswanathan, A. Vojvodic, A.T. Bell, J.K. Nørskov, The Journal of Physical Chemistry C, 116 (2012) 21077-21082.
Acknowledgements This work was financially supported by the Science Challenge Project (TZ2018004) and National Natural Science Foundation of China (51572016 and U1530401), the National Program for Thousand Young Talents of China. The computational supports from Tianhe-JK at the Beijing Computational Science Research Center (CSRC) and NSFC-Guangdong Joint Fund (the second phase) under Grant No. U1501501 are greatly acknowledged.
Pan Guo received his B. S. degree in Applied chemistry from Lanzhou University, Gansu province, China in 2013. He then joined Beijing Computational Research Center, China Academy Of Engineering
Physics. He is now a Ph. D. candidate in Prof. Li-Min Liu and Woon-Ming Lau’s Research Group. His current research interests include the electro-catalysis and Li-ion battery.
Jian Wu received his Ph.D. degree from Beijing Computational Research Center in 2016. Currently, he is working in School of Mathematics and Physics, University of Science and Technology Beijing. His present scientific interests focus on the theoretical and experimental deign of nanomaterials for energy conversion and storage.
Xi-Bo Li received his Ph.D. degree from Beijing Computational Research Center in 2016, under the supervision of Prof. Li-Min Liu. Currently, he works in Department of Physics, Jinan University. He has co-authored more than 10 journal papers. His current research is mainly focused on theoretical deign of novel materials for fuel cell and photocatalysis.
Jun Luo received his B.S. (2001) and Ph.D. (2006) from Tsinghua University, China. Then, he worked as a postdoc in Warwick University and a research fellow in Oxford University, UK. In 2011, he joined Tsinghua University as an associate professor and was granted "National Program for Thousand Young Talents of China" (namely the 1000-plan for young talents). In 2015, he moved to Tianjin University of Technology and is a full professor and the director of Center for Electron Microscopy with an aberration-corrected STEM and other five electron microscopes there. His research focuses on low-dimensional materials and electron microscopy.
W.M. Lau currently leads the Center for Green Innovation at the University of Science and Technology Beijing in China. He graduated from the Chinese University of Hong Kong (BSc) and the University of British Columbia (PhD). Professor Lau has conducted research in surface science and engineering, materials science and device engineering, nanotechnology, green energy and green manufacturing, and zero-carbon urbanization. He has published over 350 scientific articles and disclosed over 20 patents, with most patents being licensed. He holds a National Thousand-Talents Award in China and a Green Chemistry Award in Canada.
Hao Liu is an associate professor in Chengdu Green Energy and Green Manufacturing Technology R&D Center, Chengdu Development Center of Science and Technology of CAEP. He received his Ph.D. in 2010 from department of mechanical and materials engineering of the University of Western Ontario. After a short period postdoctoral research in Western Surface Science, he joined CAEP. His current research interests lie within the design of novel nanomaterials for clean energy, especially for batteries.
Xue-Liang (Andy) Sun is a Canada Research Chair in Development of Nanomaterials for Clean Energy, Fellow of the Royal Society of Canada and Canadian Academy of Engineering and Full Professor at the University of Western Ontario, Canada. Dr. Sun received his Ph.D. in materials chemistry in 1999 from the University of Manchester, UK, which he followed up by working as a postdoctoral fellow at the University of British Columbia, Canada and as a Research Associate at L’Institut National de la Recherche Scientifique (INRS), Canada. His current research interests are focused on advanced materials for electrochemical energy storage and conversion.
Li-Min Liu received his Ph.D. from Institute of Metal Research, Chinese Academy of Sciences in 2006. During his Ph.D. study, he visited Queen's University of Belfast, UK, for a year. Then he worked in Fritz-Haber-Institut (Germany), University College London (UK), and Princeton University. Since 2012, he has taken a tenure-track 26 / 27 position in Beijing Computational Science Research Center. He has co-authored more than 80 journal papers. He was granted "the 1000-plan for young talents" and "the National Science Fund for Excellent Young Scholars". His research interests focus on nanocatalysts, TiO2-based photocatalysis, electrocatalysis and fuel cells.
Highlights
3D binder-free catalyst was prepared by CVD and electrodeposition.
The Co(OH)2@NCNTs@NF possesses open porous structure.
The Co(OH)2@NCNTs@NF exhibits high activity and stability in water-splitting.
PII: DOI: Reference:
S2211-2855(18)30101-0 https://doi.org/10.1016/j.nanoen.2018.02.032 NANOEN2519
To appear in: Nano Energy Received date: 8 January 2018 Revised date: 14 February 2018 Accepted date: 15 February 2018 Cite this article as: Pan Guo, Jian Wu, Xi-Bo Li, Jun Luo, Woon-Ming Lau, Hao Liu, Xue-Liang Sun and Li-Min Liu, A highly stable bifunctional catalyst based on 3D Co(OH)2@NCNTs@NF towards overall water-splitting, Nano Energy, https://doi.org/10.1016/j.nanoen.2018.02.032 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
A highly stable bifunctional catalyst based on 3D Co(OH)2@NCNTs@NF towards overall water-splitting Pan Guo‡1,2, Jian Wu‡4, Xi-Bo Li‡1, Jun Luo5, Woon-Ming Lau4, Hao Liu2*, Xue-Liang Sun6*, Li-Min Liu1,3* 1
Beijing Computational Science Research Center, Beijing 100193, China Chengdu Green Energy and Green Manufacturing Technology R&D Center, Chengdu Development Center of Science and Technology of CAEP, Chengdu, Sichuan, 610207, China 3 School of physics, Beihang University, Beijing 100191, P. R. China 4 Center for green Innovation, School of Mathematics and Physics, University of Science and Technology Beijing, Beijing 100083, China 5 Tianjin Key Laboratory of Advanced Functional Porous Materials and Center for Electron Microscopy, School of Materials Science and Engineering, Tianjin University of Technology, Tianjin 300384, China 6 Department of Mechanical and Materials Engineering, the University of Western Ontario, London, Ontario, N6A 5B9, Canada ‡ contributed equally to this work 2
Email: [email protected] or [email protected]; [email protected]; [email protected] Abstract Electrocatalysts with low overpotential and high stability are highly demanded in water-splitting system. The efficiency of water-splitting is largely restricted by the oxygen evolution reaction (OER). Here, we developed a two-step method to prepare 3D porous material through chemical vapour deposition and electrodeposition combined with the first-principles calculations. Ultrathin α-Co(OH)2 nanosheets grown on the combined substrate of N-doped carbon nanotubes (NCNTs) and nickel foam were fabricated to investigate their electrochemical behaviour. Because of the characteristics of the ultrathin, microporous α-Co(OH)2 and its derivatives, the 3D Co(OH)2@NCNTs@NF exhibits outstanding performance as a bifunctional catalyst for water-splitting. The overpotentials to achieve 10 mA cm-2 current density in 1 M KOH for OER and hydrogen evolution reaction (HER) are 270 mV and 170 mV, respectively. The as-prepared material exhibits superior stability, which generate 10 mA cm-2 current density in overall water-splitting over 600 hours without obvious degradation in 1 M KOH at voltage of 1.72 V vs. RHE. The first-principles calculations reveal that the N-doping not only can effectively enhance the interaction
between the substrate and active material (CoOOH), but also modulate the electronic structure of CoOOH to speed up the O2 releasing during the OER. Graphical abstract:
Keywords: Water-splitting; OER; HER; Electrodeposition; N-CNT
Introduction Searching for renewable energy productions is a rather critical issue to reduce the dependence on the fossil fuels[1-3]. Electrocatalytic water-splitting is one of the most promising ways to produce clean energy. In order to meet the requirements of the applications, it is long desired to develop cheap, non-toxic and efficient catalysts to stimulate the slow kinetics of water-splitting[4, 5]. The water-splitting process contains both OER and HER[6-8], and OER is much harsher than HER due to the complex four electrons process, which greatly restrains the efficiency and kinetics of the whole reaction[9, 10]. On the other side, some highly efficient OER catalysts do not exhibit good performance in HER because of the different
working medium. Therefore, developing highly-efficient, non-precious bifunctional catalysts which can be used in the same electrolyte is the ultimate way to meet industry demand. Cobalt is one of the earth-abundant elements and its corresponding hydroxide material is very easy to synthesize. Cobalt hydroxides are known to exist in two polymorphs, namely α- and β-Co(OH)2. Among them, the α-Co(OH)2 exhibits higher activity due to the fact that the layer-layer distance of α-Co(OH)2 is larger than that of β-Co(OH)2. However, most of the Co(OH)2 prepared are β-Co(OH)2[11, 12], so it is necessary to investigate the performance of α-Co(OH)2 as an electrocatalyst to catalyse water-splitting. Firstly, the NCNTs was deposited on the nickel foam (NF), then the Co(OH)2 nanosheets were grown on the NCNTs to form a binder-free 3D porous catalyst. This unique design has the following advantages: enhancing the accessible effective surface areas of the catalyst, facilitating escape of gases, improving the electron transfer and effectively avoiding aggregation/agglomeration during OER process[13, 14]. To the best of our knowledge, this work is the first systematic report focuses on α-Co(OH)2 to investigate its catalytic performance in water-splitting. Based on the aforementioned facts, a novel 3D self-supported porous material, namely Co(OH)2@NCNTs@NF, was fabricated, which exhibits a small overpotential of 270 mV in OER at 10 mA cm-2 current density in 1 M KOH over 200 h without any degradation. Meanwhile excellent catalytic performance towards HER reaction was obtained, which needs 170 mV overpotential to generate 10 mA cm-2 current density in 1 M KOH. We also constructed a two electrodes system to investigate catalytic performance in overall water-splitting and our catalyst generate 10 mA cm-2 current density in 1 M KOH with only 490 mV overpotential needed and exhibit such an amazing stability in 10 mA cm-2 water-splitting process (>600 h).
Experiment results SEM and TEM images show that the NCNTs were uniformly distributed over the entire nickel foam (Fig.
S1a). The nanotubes intertwined with each other and connected to the nickel foam directly, forming a binder-free 3D porous skeleton (Fig. S1b). From the magnified image (Fig. S1c and S1d), the NCNTs exhibits a bamboo-like tubular structure and a rough surface, which are the typical characteristics of NCNTs[15-17]. This 3D structure can facilitate fast transportation of electrons between NF and NCNTs as well as effective mass exchange in electrolyte, and rough surface of NCNTs are beneficial for electrodepositing. Following this, the Co(OH)2 nanosheets were evenly deposited onto the surface of each nanotubes, forming a blue thin film on the surface of black NCNTs (Fig. S2). The SEM images (Fig. 1a and S3) show that the Co(OH)2 nanosheets interconnected with each other to form a nanosheet array structure. The low magnification SEM image presented in Fig. 1b shows that the Co(OH)2 nanosheets completely cover over the NCNTs. As a comparison, Co(OH)2 nanosheets were also electrodeposited onto bare NF under the same experimental conditions. As shown in Fig. 1c, the nanosheets on NCNTs and NF are greatly different. The thickness and lateral size of the nanosheets on NF are obviously thicker and smaller than those on NCNTs, respectively, which clearly verities the importance of NCNTs for loading materials[18-20]. The TEM image (Fig. 1d) further verifies the intimate and direct contact between Co(OH)2 and NCNTs. The TEM image (Fig. S4) of the marked area in Fig. 1d exhibits that the thickness of the nanosheet is around 6 nm. As displayed in TEM image of Fig. 1e, the nanosheets are nearly transparent, indicating their ultrathin feature. The spacings of (002) and (100) atomic planes are determined to be 0.46 nm and 0.27 nm (Fig. 1f), respectively, which is consistent with the data of α-Co(OH)2[21]. Meanwhile, the XRD result further suggests that the electrodeposited material is α-Co(OH)2 (see in Fig. S5a). The selected-area electron diffraction pattern shown in inset of Fig. 1f exhibits a diffused spot pattern, further confirming the polycrystalline nature of the electrodeposited Co(OH)2. The elemental mapping of Co(OH)2@NCNT (Fig. 1g) shows a uniform distribution of C, Co, O and N elements in the composite. Among them, C and N elements exist in the region of NCNT, while Co and O elements distributed over
the entire region, which again verifies the designed structure.
Fig. 1 | The morphology characterization of the as-prepared material (a) SEM image of Co(OH)2 nanosheets on NCNTs. (b) low magnification SEM image of Co(OH)2@NCNTs. (c) The SEM image of Co(OH)2 electrodeposited on NF. (d) TEM image of Co(OH)2@NCNTs. (e) TEM image of Co(OH)2 nanosheets. (f) HRTEM image of Co(OH)2 nanosheets (inset: SAED pattern of the nanosheets). (g) Elemental mapping of C, N, Co and N in Co(OH)2@NCNTs composite.
In order to obtain a more stereoscopic and realistic image of the Co(OH)2@NCNTs, 3D reconstruction method were taken to restore the original 3D structure of the as-prepared material. The original image (Fig. 2a) exhibits clear stereoscopic structure of Co(OH)2 nanosheets. The nanosheets grow from NCNT axis and have a rough surface which exposes more active area than smooth surface. Meanwhile, the nanosheets are not stacked to each other, thus the open spaces between nanosheets ensure enough contact between electrolyte and Co(OH)2 surface. Fig. 2b displays the transparent mode of the original 3D structure, in which the CNT, doped-N and Co(OH)2 nanosheets are clearly seen. The
N-doped CNTs possess concave-convex structure (seen in Fig. S6) which is beneficial for robust adhesion between Co(OH)2 nanosheets and NCNTs. The top view image (shown in Fig. 2c) of the truncated section in Fig. 2b implies that the nanosheets grow radially from the NCNT axis and the emerged areas between nanosheets and N may be the original nucleation sites.
Fig. 2 | The 3D reconstruction characterization of the Co(OH)2 nanosheets on N doped CNT.
The chemical composition of Co(OH)2@NCNT@NF is further determined by X-ray photoelectron spectroscopy (XPS). The results are shown in Fig. 3a and b. There are two peaks at 781.4 and 797.3 eV accompanied by two satellite peaks at 786.0 and 803.0 eV in the Co 2p spectrum (Fig. 3a), indicating the oxidation state of Co is +2. The XPS spectrum of O 1s (Fig. 3b) exhibits single peak at 531.2 eV, corresponding to the bound hydroxide groups[22]. These results collectively confirm the synthesis of Co(OH)2. The XPS spectrum of N 1s (Fig. S7) can be deconvoluted into pyridine N (398.6 eV), oxidized N (402.3 eV) and graphite N (401.0 eV), among which pyridinc and graphitic N are considered as the OER active sites[23].
Fig. 3 | The XPS spectra of prepared material (a) Co 2p and (b) O 1s XPS spectra of Co(OH)2@NCNTs@NF before being tested. (c) Co 2p and (d) O 1s XPS spectra of Co(OH)2@NCNTs@NF after stability test.
Electrochemical Results A series of electrochemical experiments were carried out in 1 M KOH to investigate the performance of the as-prepared catalysts. We made three batches of Co(OH)2@NCNTs@NF with different electrodeposition time to determine the suitable time. Taking the onset potential and overpotential at 10 mA cm-2 as judging criterion, the 20 minutes sample shows the best performance in both onset potential and overpotential (Fig. S8), so in the following experiment 20 minutes was taken as the electrodeposition time. At first, the linear sweep voltammetry data of Co(OH)2@NCNTs@NF, Co(OH)2@NF, NCNTs@NF and NF were recorded. As shown in Fig. 4a, the substrate NF shows a negligible current in the voltage range from 1.40 V to 1.60 V. Thus, it is a suitable substrate for studying the performance of high active materials in low overpotential ranges. Co(OH)2@NF and NCNTs@NF also exhibit relatively unsatisfactory catalytic activity in this voltage range. The as-prepared catalyst, Co(OH)2@NCNTs@NF, manifests superior advantages not only in onset potential but also in OER current at any given voltages. Besides, the current density of
Co(OH)2@NCNTs@NF at 1.50 V is higher than the sum of both NCNTs@NF and Co(OH)2@NF at the same voltage. Therefore, a synergistic effect must exist between Co(OH)2 and NCNTs to promote the catalytic activity of Co(OH)2@NCNTs@NF. The LSV data of Co(OH)2@NCNTs@NF exhibits two peaks at 1.10 V and 1.43 V[12, 24, 25], which corresponds to the transformation of Co2+/Co3+ and Co3+/Co4+, respectively. The formation of Co4+ is critical for improving the catalytic activity of Cobalt-based materials because the Co4+ is considered as the main active sites for OER[26-28]. As shown in the LSV data of Co(OH)2@NF, one peak exists at 1.13 V, while the peak corresponding to transformation of Co3+/Co4+ is negligible, which is due to the competition reactions[29]. The negligible peak of transformation of Co3+/Co4+ indicates that little Co4+ formed in Co(OH)2@NF comparing to Co(OH)2@NCNTs@NF, consequently, the performance of Co(OH)2@NF is relatively unsatisfactory. As shown in the inset of Fig. 4a, the capacitance reaction is completed before the OER, therefore the current after 1.47 V mainly comes from OER. The as-prepared Co(OH)2@NCNTs@NF only needs 270 mV overpotential to generate 10 mA cm-2 current density, which is among the limited materials with overpotential lower than 300 mV to generate 10 mA cm-2 current density[14,
30-34]
. Such a small
overpotential is the lowest one according to our knowledge in either Co or Co(OH) 2-based materials. The performance in high current density conditions is also excellent. As shown in Fig. 4a, it can reach 100 mA cm-2 current density at the overpotential of 410 mV, which is superior in current cobalt-based catalysts. As shown in Fig. 4b, the Tafel curves of the three materials are plotted[30], and the Tafel slope of Co(OH)2@NCNTs@NF is clearly lower than those of Co(OH)2@NF and NCNTs@NF. Considering that lower Tafel slope means faster catalytic kinetics[31, 35, 36]; thus, Co(OH)2@NCNTs@NF exhibits a better catalytic performance than their counterparts. Although the Tafel slope of Co(OH)2@NCNTs@NF is not comparable to the state-of-the-art precious metal catalysts, the relatively low value of 72 mV/dec also shows advantage in transition metal-based catalysts. Long time duration of the catalysts in the working life span is another important parameter. The
overall water-splitting is usually carried out at a constant current or voltage. In order to mimic a more realistic working condition, the chronopotentiometry was used to evaluate the catalyst stability. The chronopotentiometry test was taken at 10 mA cm-2, Co(OH)2@NCNTs@NF exhibits an amazing stability over 200 hours without any degradation (see Fig. 4c). In order to mimic a harsher condition, another stability test was taken. At this test, the sample was firstly tested for 100 hours at 10 mA cm-2, then for another 100 hours test at 20 mA cm-2. The Co(OH)2@NCNTs@NF maintains a nearly constant voltage for the final 100 hours at 20 mA cm-2 (Fig. 4d). Such a long and steady stability performance excludes the corrosion of carbon nanotubes and also verifies the strong structure of our catalyst[33]. To take a more detailed comparison between the catalyst before and after testing, the LSV data of the catalyst after testing is recorded. The comparison data were shown in Fig. 4e, the LSV curve of the catalyst after testing at 10 mA cm-2 for 100 hours is similar to that after another 100 hours at 20 mA cm-2 except a little degradation for the latter one at voltage > 1.60 V. The efficient HER catalysts usually work at acid medium, which greatly limits the application of transition metal oxides and hydroxides, and thus, it is rather challenging to combine OER and HER catalysts in one electrolyte medium[32, 34]. Therefore, efficient bifunctional catalysts working in one medium are extremely desired. The HER performance of Co(OH)2@NCNTs@NF is also examined in 1 M KOH. The optimal electrodeposition time for catalyst used in HER is determined to be 15 minutes (see Fig. S9). The LSV curve presented in Fig. 4f clearly demonstrates the advantages of Co(OH)2@NCNTs@NF over their counterparts. The Co(OH)2@NCNTs@NF electrode reaches 10 mA cm-2 current density at a small overpotential of 170 mV. The chronopotentiometry measurement was also taken to investigate its stability. As shown in Fig. S10, the Co(OH)2@NCNTs@NF electrode undergoes an activation process during stability test. The overpotential needed to generate 10 mA cm-2 current density gradually decreased to 115 mV after 120 h testing and was stable around this potential during the following 400 h testing.
Owing to the excellent catalytic performance, the Co(OH)2@NCNTs@NF can be used as OER and HER electrodes to catalyse overall water-splitting. As shown in Fig. S11, Co(OH)2@NCNTs@NF exhibits an excellent performance with a voltage of 1.72 V needed to generate 10 mA cm -2 current density in overall water-splitting. Chronopotentiometry test was further carried out to measure the durability of the as-prepared catalyst. As presented in Fig. 4g, the Co(OH)2@NCNTs@NF electrode exhibits the good stability over the whole testing period (600 h), clearly indicating its attractiveness in practical applications.
Fig. 4 | Electrochemical characterizations. (a) Linear sweep voltammetry curves of the Co(OH)2@NCNTs@NF electrode and their counterparts for OER in 1 M KOH solution at 2 mV s-1. (b) Tafel plots of the three electrodes in 1 M KOH solution at 2 mV s-1. (c) Chronopotentiometry curves obtained with the Co(OH)2@NCNTs@NF act as OER electrode in 1 M KOH at constant current density of 10 mA cm-2. (d) Chronopotentiometry curves taken at 20 mA cm-2 in 1 M KOH after the Co(OH)2@NCNTs@NF electrode being tested in 1 M KOH for 100 h at 10 mA cm-2. (e) Linear sweep voltammetry curves of the Co(OH)2@NCNTs@NF electrode after Chronopotentiometry test. (f) Linear sweep voltammetry curves of
the Co(OH)2@NCNTs@NF electrode and their counterparts for HER in 1 M KOH solution at 2 mV s-1. (g) Chronopotentiometry curves taken at 10 mA cm-2 in 1 M KOH toward overall water-splitting.
XRD and XPS were carried out on Co(OH)2 after cycling, and interestingly, the results show that the α-Co(OH)2 transformed into ϒ-CoOOH after cycling ( Fig. 3c and 3d). As shown in Fig. S5b, the XRD diffraction pattern belongs to ϒ-CoOOH[37]. Although there is no specific PDF card in JCPDS corresponding to ϒ-CoOOH, the peaks at about 13.66°, 26.98°, 37.88° and 39.73° approximately match the (003), (006), (101) and (105) peaks of ϒ-NiOOH[38, 39]. As shown in Fig. 3c, the XPS spectrum of Co after cycling is different from that of the pristine one. The best deconvolution of Co 2p can be accomplished by the assumption that Co2+ and Co3+ coexist in the material. As shown in Fig. 3c, the dominant peaks at 780.0 and 795.0 eV belong to the Co3+ 2p3/2 and 2p1/2, respectively. The two satellite peaks at 790.0 and 805.0 eV further indicates the formation of CoOOH during OER process. Also, the XPS spectrum of O 1s (Fig. 3d) is different from that of the original one with two peaks located at 530.9 and 529.1 eV, which are attributed to the O from the OH and oxide ions[40]. The XPS spectrum of O 1s in CoOOH is different from that of either Co(OH)2 or Co3O4, and these latter two materials have the main peak located in 531.2 and 530.0 eV, respectively[40]. The difference of O 1s spectrum in CoOOH with those in Co(OH)2 and Co3O4 clearly confirms the formation of CoOOH.
Discussion First of all, the ultrathin Co(OH)2 nanosheets contain many micropores (Fig. 5a) which can expose more grain boundaries to participate in OER. As discussed above, the α-Co(OH)2 is different from the widely explored β-Co(OH)2, whose interlayer distance is about twice larger. The relatively large interlayer distance provides more active sites and the thin nanosheets shorten the diffusion length of electrolyte. Meanwhile, this property is inherited by its derivative, ϒ-CoOOH, which also exhibits the relatively large
interlayer distance. Besides that, the surface CoOOH formed during OER process on the Co(OH) 2 nanosheets is a better conductor for electrons[41, 42]. Secondly, the unique spatial structure plays a key role in catalysing OER. As shown in SEM image in Fig. S1b, the as-prepared catalyst featured a woven-like intertwined 3D network structure, which are beneficial for fast transportation of electrolyte and electrons. What’s more, the tubular structure and rough surface, typical characters of NCNTs, can provide more nucleate sites for uniformly electrodepositing Co(OH)2. The interaction between NCNTs and Co(OH)2 is quite strong and no visible voids exist between NCNTs and Co(OH)2 (Fig. 1d). This intimate contact is very important to retain a robust structure and strong synergistic effect between N and Co. It is worth mentioning that it takes 20 minutes ultrasonication to disperse the catalyst during TEM sample preparation, which again confirms the robust structure of the catalyst. The structure after the stability test were also examined. As shown in Fig. 5b and 5c, SEM and TEM images indicate that the catalyst maintains its original structure after 200 hours testing.
Fig. 5 | The morphology characterization of the Co(OH)2@NCNTs. (a) Magnified TEM image of Co(OH)2 nanosheets. (b) SEM and (c) TEM images of the Co(OH)2@NCNTs@NF electrode after stability test at 10 mA cm-2 for 200 h.
Last, the role of N dopant is also very important. As analysed in XPS section, there are three kinds of N, namely pyridine N, oxidized N and graphite N, among which pyridinic and graphitic N atoms are considered to be the active sites for OER reaction[23]. To understand why the N-doping can greatly improve the performance of the CoOOH in OER, first-principles calculations were carried out. First of all, the most stable configurations of monolayer β-CoOOH and the heterostructure between monolayer
β-CoOOH and graphene with one pyridinic N defect are examined to mimic the interaction between the CoOOH and NCNT. The relative stability of several different monolayer CoOOH configurations are calculated. Four typical monolayer CoOOH configurations are shown in Fig. S13. The most stable configuration of monolayer CoOOH (denoted as CoOOH@) is shown in Fig. S13(a), and the main difference from others is that H atoms are distributed equally on the oxygen atoms. Based on the most stable configuration of monolayer CoOOH, the heterostructure between monolayer CoOOH and graphene with the pyridinic N defect are constructed. The corresponding configurations and the relative energies are shown in Fig. S13(e) – (f). The heterostructure in Fig. S13(e) is the most stable one, in which the N atom bonds with the cobalt atoms in CoOOH (denoted as CoOOH@N1). Based on the most stable configuration of CoOOH@ and CoOOH@N, the OER processes were calculated. In order to know the effect of N concentration on the OER, the system containing two N atoms was also considered (denoted as CoOOH@N2, Fig. S13(h)). Our calculation results show that the formation energies of oxygen vacancy in CoOOH@, CoOOH@N and CoOOH@N2 are 2.56, 2.19 and 0.19 eV per oxygen vacancy, respectively. This trend indicates that the oxygen vacancy is relatively easier to form in heterostructure with one N defect than the one in the pure monolayer CoOOH, and even easier in heterostructure with more N defects. In the following, the OER activities of CoOOH@, CoOOH@N1 and CoOOH@N2 are calculated. Fig. 6 shows the free energy changes of OER steps on CoOOH@, CoOOH@N1 and CoOOH@N2, and the corresponding intermediate species of OER are shown in Fig. S14 and Fig. S15. The first principles calculations indicate that the last step of O2 formation (*OOH + → *O2(g) + H+ + e-) is the potential determining step (PDS) on CoOOH@ and CoOOH@N1, with the free energy barriers of 2.83 eV and 0.56 eV at U = 1.23 eV vs SHE, respectively. The reduced overpotential for heterostructure exhibits that CoOOH@N1 show better OER performance than the pure CoOOH@. Meanwhile, with the increase of the concentration of N, the overpotential of OER in CoOOH@N2 become 0 eV. These results clearly
suggest that the N dopant can greatly enhance OER activity.
Fig. 6 | Free energy changes diagram at pH = 13, T = 298 K for the four steps of OER at U= 0 eV vs SHE (black solid line) at equilibrium potential U = 1.23 eV vs SHE (red dashed line) on (a) CoOOH@, monolayer CoOOH; (b) CoOOH@N1, hetrostrucuture of CoOOH binding with one pyridinic N defect; (c) CoOOH@N2: hetrostrucuture of CoOOH binding with two pyridinic N defects. The overpotential are in parentheses and the negative value in the (c) indicates the overpotential is 0 eV.
In order to unveil the origin of the high OER activity of the heterostructure, the electron transfers from the monolayer CoOOH to the graphene were calculated. For the CoOOH@, there is no electron transfer considering the system does not contain the interface. While, the corresponding values are 0.16 e for CoOOH@N1 and 0.58 e for CoOOH@N2, respectively. In addition, the charge density difference of CoOOH@N1 is also given to show the distribution of the electron. As show in Fig. S16, it clearly shows that there is electron transfer from Co atoms in CoOOH_Ov to N atom in pyridinic N defect in graphene. As shown in Fig. 6, with increase of the electron transfer, the free energy of adsorption of *OOH on CoOOH, △G(OOH), becomes weaker: 0.09 eV for CoOOH@, 2.36 eV for CoOOH@N1, 3.69 eV for CoOOH@N2, which facilitates the O2 formation step (*OOH + → *O2(g) + H+ + e-). Meanwhile, the adsorption ability of *O is correlated with the formation energy of oxygen vacancy for CoOOH@, CoOOH@N, and CoOOH@N2. The smaller the formation energy of oxygen vacancy in CoOOH, the smaller △G(O) for the corresponding OER. Further, the OER activities of CoOOH@, CoOOH@N, and CoOOH@N2 are greatly related with the
oxidation state of the Co atoms around the oxygen vacancy. The total charge state of three Co atoms nearby the oxygen vacancy in CoOOH@, CoOOH@N, CoOOH@N2 are +1.76 e, +2.25 e, and +2.41 e, respectively. As shown in Fig. 7, CoOOH@N and CoOOH@N2 loses more electrons to pyridinic N defect than CoOOH@, the Co atoms in CoOOH@N and CoOOH@N2 behave higher charge state, which greatly decrease the ability of strongly adsorbing O or OOH by transferring the extra electron to O or OOH. In all, the electrons transfer in the heterostructure makes the cobalt atoms around oxygen vacancy behave as the high oxide state, effectively decreasing the adsorption ability of *OOH and the overpotential of OER.
Fig. 7 | The relationship between electron transfer in the heterostructure and △G(O) (blue square), △G(OOH) (red round) and Gf(Ov) (black triangle) for CoOOH@, CoOOH@N, and CoOOH@N2.
Conclusions Owing to the merits of nickel foam and N-doped carbon nanotubes, we take a combined method including CVD and electrodeposition to prepare 3D porous catalyst Co(OH)2@NCNTs@NF, which is confirmed to be an excellent catalyst with small overpotential and long stability. This material can generate OER current density of 10 mA cm-2 at voltage of 1.50 V vs. RHE, corresponding to an
overpotential of 270 mV, which is among the limited materials with an overpotential lower than 300 mV to reach 10 mA cm-2 current density. The stability of our catalyst is also significantly improved with a working life span lasting over 200 hours without degradation. This excellent material also catalyse HER with good performance. It only needs 170 mV overpotential to generate 10 mA cm -2 current density in 1 M KOH. Benefitting from the excellent performance towards OER and HER, our material catalyse overall water-splitting at 1.72 V vs. RHE to generate 10 mA cm-2 current density and this performance can maintain over 600 hours without degradation. Therefore our preparation method can be expected to be adopted in industry application.
Experimental section Preparation of N-doped CNTs on nickel foam We use a chemical vapor deposition (CVD) method developed in our lab to prepare N-doped CNTs on nickel foam[43-45]. The original nickel foam will become brittle after annealing, which is unfavorable for acting as a substrate material. After a series of trial and error, we found that the pressed nickel foam can sustain its ductility after annealing. Thus, all the nickel foam were pressed before using. Before the deposition process, the nickel foam were cut into 1 cm × 1.5 cm squares and pressed to 0.16 mm thickness, and then put into a 6 M HCl solution to sonicate for 30 minutes. After that the nickel foam were rinsed with copious deionized water (DI) and ethanol to remove the residue impurity. The obtained nickel foam were put in an electronic vacuum oven for 12 hours to vapor the trace amount of liquid. Following that, the nickel foam were used as the substrate to load N-doped CNTs. Briefly, the nickel foam were placed in the middle of a quartz tube. The system was purged with high purity argon gas with a flow rate of 800 sccm for 30 minutes to ensure an inert environment. After that the furnace was heated in a programmable step to 900 C in 30 minutes and then kept at this temperature for another 15 minutes. Ferrocene, ethylene gas and melamine which acts as the catalyst, carbon source and nitrogen
source, respectively, was introduced into the tube lasting for 15 minutes to deposition N-doped CNTs. Finally after turning down the ethylene gas, the furnace was left to cool down to room temperature, during which the argon gas was continue purged to protect the obtained material. The obtained material was donated as NCNTs@NF. Preparation of Co(OH)2@NCNTs@NF The electrodeposition process was completed in a standard three electrode system with the NCNTs@NF as the working electrode, platinum mesh as the counter electrode and saturated calomel electrode (SCE) as the reference electrode. The Co(OH)2 nanosheets were electrodeposited onto the NCNTs@NF substrate in an aqueous solution containing 6 mmol Co(NO3)2•6H2O at a constant cathodic current density of 1.5 mA cm-2 for 20 minutes. Following that, the obtained Co(OH)2@NCNTs@NF was washed with copious water and ethanol for several times repeatedly. Finally, the as-prepared sample was vacuum dried at 60 C for 2 hours. The loading mass of Co(OH)2 nanosheets was determined by the mass difference between the substrate before and after electrodeposition using a high precise electronic balance. The average mass was around 0.72 mg cm-2. We also electrodeposited Co(OH)2 onto bare nickel foam in the same condition except for the substrate as a comparison to study the possible synergistic effect between Co(OH)2 and N-doped CNT.
Material characterization The phase purity and crystallographic information were obtained by using the X-ray diffraction (XRD, D/max 2200/PC, Rigaku, 40 kV, 20 mA, Cu K α radiation, λ = 1.5406 Å). The surface morphology and elemental composition of the sample were investigated by scanning electron microscopy (SEM, Hitachi S-5200). The more detailed microstructure was characterized by using high resolution transmission electron microscope (HRTEM, FEI, Tecnai F20) equipped with energy dispersive X-ray spectroscopy elemental mapping. The 3D reconstruction were carried out by Transmission Electron Microscope with A
Probe Corrector (Titan Cubed Themis G2 300).
Electrochemical measurements All the involved electrochemical measurements were carried out in a CHI 760E electrochemical workstation in 1 M KOH aqueous solution at room temperature. The prepared Co(OH)2@NCNTs@NF, Co(OH)2@NF and NCNTs@NF were used as the working electrodes directly without any further treatment. A platinum mesh and saturated calomel electrode was used as the counter and reference electrode, respectively. The SCE was calibrated using a method described in Dai’s work[46] and the relationship between RHE and SCE in 1 M KOH is determined to be: E RHE = ESCE + 1.062 V (see in Fig. S12). Before any measurements the three electrode system was purged with high purity O 2 gas for 20 minutes and O2 was maintained during the whole testing to ensure the equilibrium of O2/H2O at 1.23 vs. RHE[30]. The electrochemical data were recorded only after the system was cycled to reach a stable cyclic voltammogram. The cyclic voltammogram (CV) and linear sweep voltammetry (LSV) was performed at a scan rate of 50 mV s-1 and 2 mV s-1, respectively. The scan rate choice of LSV is slow enough to minimize the capacitance current and obtain accurate polarized OER current curve. All the obtained results are not compensated by iR drop because the internal resistance is very small. Chronopotentiometry measurement was carried out in the same experimental setup with no iR drop compensation.
First-principles calculations. All calculations were performed using DFT methods, as implemented in the CP2K/Quickstep package[47]. The exchange correlation potential was described by the generalized-gradient approximation (GGA) with the spin-polarized functional of Perdew-Burke-Ernzerh (PBE)[48]. A cutoff energy of 500 Ry was chose for the wave functions. Core electrons were modeled by norm-conserving pseudopotentials[49] with 17, 6, 1, 4 and 5 valence electrons of Co, O, H, C and N, respectively. Only the Γ-point was used for Brillouin zone integration. The DFT+U method, based on the
Mullikan 4d state population analysis, was used to describe the Co 4d electrons. A U value of 3.52 eV for Co was used in all calculations[50]. Besides, van der Walls interaction was also considered with the D2 [51] in our DFT calculations. The details of the models used are as discussed as follows. In the monolayer CoOOH models, the (2×3) supercell were explored for further calculations. In the heterostructure model, the (2×3) monolayer CoOOH and the (4√2×5√2)R30。graphene with pyridinic N defects are perpendicular to each other. In order to explore effect of the density of pridinic defect on OER, the heterostructure models with one and two pridinic defects in graphene of the heterstructure were established. In additional, there was one oxygen vacancy in both in monolayer CoOOH model and heterostructure models.
Free Energy calculations. As proposed in the previous work, it is more convenient to model the thermochemistry of OER at acid condition[50, 52], where the steps involving electron transfer in step (1)─(4) are modified as follow *+ H2O +→OH*+H++ e-
(1)
OH*→O *+ H++e-
(2)
O*+ H2O →OOH*+ H++e-
(3)
OOH*+→*+O2(g)+ H++e-
(4)
where * stands for the activity site on the catalyst’s surface. The change of Gibbs free energy in each reaction step could be expressed as: △G = △E + △ZPE + △H - T△S+△GH+(pH), where △E, △ZPE, △H, T△S , △GH+(pH)indicate the changes of electronic energy, zero point energy, the enthalpy, the entropy, free energy change of proton relative to pH = 0, respectively. And more pH effect on H+ is considered as △GH+(pH) = -kBT ln(10) × pH[53]. In our calculations, the zero point energy was not considered. Using the standard hydrogen electrode (T = 298.15 K, P = 1 bar, pH = 0), the Gibbs energy of the proton/electron could be expressed as that of H2, GH++ Ge-=1/2 GH2. As the difficulty of determining
the energy of O2 within GGA-DFT frame, the free energy of O2 could be induced by equation: GO2(g) = 2 GH2O - 2 GH2 + 4.92 eV.
References [1] M. Lefevre, E. Proietti, F. Jaouen, J.P. Dodelet, Science, 324 (2009) 71-74. [2] C. Liu, F. Li, L.P. Ma, H.M. Cheng, Adv Mater, 22 (2010) E28-62. [3] K.S. Joya, Y.F. Joya, K. Ocakoglu, R. van de Krol, Angew Chem Int Ed Engl, 52 (2013) 10426-10437. [4] A. Aijaz, J. Masa, C. Rösler, W. Xia, P. Weide, A.J.R. Botz, R.A. Fischer, W. Schuhmann, M. Muhler, Angewandte Chemie International Edition, 55 (2016) 4087-4091. [5] N.S. Lewis, D.G. Nocera, Proceedings of the National Academy of Sciences, 103 (2006) 15729-15735. [6] M.G. Walter, E.L. Warren, J.R. McKone, S.W. Boettcher, Q. Mi, E.A. Santori, N.S. Lewis, Chem Rev, 110 (2010) 6446-6473. [7] C.C. McCrory, S. Jung, I.M. Ferrer, S.M. Chatman, J.C. Peters, T.F. Jaramillo, J Am Chem Soc, 137 (2015) 4347-4357. [8] F. Song, X. Hu, Nat Commun, 5 (2014) 4477. [9] R.D.L. Smith, M.S. Prévot, R.D. Fagan, Z. Zhang, P.A. Sedach, M.K.J. Siu, S. Trudel, C.P. Berlinguette, Science, 340 (2013) 60-63. [10] T.R. Cook, D.K. Dogutan, S.Y. Reece, Y. Surendranath, T.S. Teets, D.G. Nocera, Chem Rev, 110 (2010) 6474-6502. [11] J.-X. Feng, L.-X. Ding, S.-H. Ye, X.-J. He, H. Xu, Y.-X. Tong, G.-R. Li, Advanced Materials, 27 (2015) 7051-7057. [12] J.A. Koza, C.M. Hull, Y.-C. Liu, J.A. Switzer, Chemistry of Materials, 25 (2013) 1922-1926. [13] X. Yu, M. Zhang, J. Chen, Y. Li, G. Shi, Advanced Energy Materials, 6 (2016) 1501492-1501501. [14] X. Lu, C. Zhao, Nat Commun, 6 (2015) 6616.
[15] X. Xu, S. Jiang, Z. Hu, S. Liu, ACS Nano, 4 (2010) 4292-4298. [16] Z. Chen, D. Higgins, Z. Chen, Carbon, 48 (2010) 3057-3065. [17] D. Geng, H. Liu, Y. Chen, R. Li, X. Sun, S. Ye, S. Knights, Journal of Power Sources, 196 (2011) 1795-1801. [18] Y. Liang, Y. Li, H. Wang, H. Dai, J Am Chem Soc, 135 (2013) 2013-2036. [19] H. Zhu, J. Zhang, R. Yanzhang, M. Du, Q. Wang, G. Gao, J. Wu, G. Wu, M. Zhang, B. Liu, J. Yao, X. Zhang, Adv Mater, 27 (2015) 4752-4759. [20] M.-R. Gao, Y.-F. Xu, J. Jiang, Y.-R. Zheng, S.-H. Yu, J Am Chem Soc, 134 (2012) 2930-2933. [21] Z. Liu, R. Ma, M. Osada, K. Takada, T. Sasaki, J Am Chem Soc, 127 (2005) 13869-13874. [22] H.B. Li, M.H. Yu, X.H. Lu, P. Liu, Y. Liang, J. Xiao, Y.X. Tong, G.W. Yang, ACS Appl Mater Interfaces, 6 (2014) 745-749. [23] Y. Zhao, R. Nakamura, K. Kamiya, S. Nakanishi, K. Hashimoto, Nat Commun, 4 (2013) 2390. [24] M.A. Sayeed, T. Herd, A.P. O'Mullane, Journal of Materials Chemistry A, 4 (2016) 991-999. [25] Y.-C. Liu, J.A. Koza, J.A. Switzer, Electrochimica Acta, 140 (2014) 359-365. [26] B.E. Conway, T.C. Liu, Materials Chemistry and Physics, 22 (1989) 163-182. [27] G.W. Simmons, E. Kellerman, H. Leidheiser, Journal of The Electrochemical Society, 123 (1976) 1276-1284. [28] G.W. Simmons, A. Vértes, M.L. Varsányi, H. Leidheiser, Journal of The Electrochemical Society, 126 (1979) 187-189. [29] J.A. Koza, C.M. Hull, Y.-C. Liu, J.A. Switzer, Chemistry of Materials, 25 (2013) 1922-1926. [30] S. Chen, J. Duan, P. Bian, Y. Tang, R. Zheng, S.-Z. Qiao, Advanced Energy Materials, 5 (2015) 1500936-1500943. [31] X. Long, J. Li, S. Xiao, K. Yan, Z. Wang, H. Chen, S. Yang, Angew Chem Int Ed Engl, 53 (2014) 7584-7588.
[32] H. Wang, H.W. Lee, Y. Deng, Z. Lu, P.C. Hsu, Y. Liu, D. Lin, Y. Cui, Nat Commun, 6 (2015) 7261. [33] M. Gong, Y. Li, H. Wang, Y. Liang, J.Z. Wu, J. Zhou, J. Wang, T. Regier, F. Wei, H. Dai, J Am Chem Soc, 135 (2013) 8452-8455. [34] H. Jin, J. Wang, D. Su, Z. Wei, Z. Pang, Y. Wang, J Am Chem Soc, 137 (2015) 2688-2694. [35] X. Li, Y. Fang, X. Lin, M. Tian, X. An, Y. Fu, R. Li, J. Jin, J. Ma, Journal of Materials Chemistry A, 3 (2015) 17392-17402. [36] S. Liu, L. Li, H.S. Ahn, A. Manthiram, Journal of Materials Chemistry A, 3 (2015) 11615-11623. [37] Z. Chang, H. Li, H. Tang, X.Z. Yuan, H. Wang, International Journal of Hydrogen Energy, 34 (2009) 2435-2439. [38] X.-Z. Fu, Y.-J. Zhu, Q.-C. Xu, J. Li, J.-H. Pan, J.-Q. Xu, J.-D. Lin, D.-W. Liao, Solid State Ionics, 178 (2007) 987-993. [39] L. Jun, L. Rong, W. Jianming, S. Hang, Journal of Power Sources, 79 (1999) 86-90. [40] J. Yang, H. Liu, W.N. Martens, R.L. Frost, The Journal of Physical Chemistry C, 114 (2010) 111-119. [41] V. Pralong, A. Delahaye-Vidal, B. Beaudoin, J.B. Leriche, J.M. Tarascon, Journal of The Electrochemical Society, 147 (2000) 1306. [42] J. Huang, J. Chen, T. Yao, J. He, S. Jiang, Z. Sun, Q. Liu, W. Cheng, F. Hu, Y. Jiang, Z. Pan, S. Wei, Angew Chem Int Ed Engl, 54 (2015) 8722-8727. [43] H. Liu, Y. Zhang, R. Li, X. Sun, S. Désilets, H. Abou-Rachid, M. Jaidann, L.-S. Lussier, Carbon, 48 (2010) 1498-1507. [44] R. Mi, S. Li, X. Liu, L. Liu, Y. Li, J. Mei, Y. Chen, H. Liu, H. Wang, H. Yan, W.-M. Lau, Journal of Materials Chemistry A, 2 (2014) 18746-18753. [45] J. Wu, P. Guo, R. Mi, X. Liu, H. Zhang, J. Mei, H. Liu, W.-M. Lau, L.-M. Liu, Journal of Materials Chemistry A, 3 (2015) 15331-15338. [46] Y. Liang, Y. Li, H. Wang, J. Zhou, J. Wang, T. Regier, H. Dai, Nat Mater, 10 (2011) 780-786.
[47] J. VandeVondele, M. Krack, F. Mohamed, M. Parrinello, T. Chassaing, J. Hutter, Computer Physics Communications, 167 (2005) 103-128. [48] J.P. Perdew, A. Ruzsinszky, G.I. Csonka, O.A. Vydrov, G.E. Scuseria, L.A. Constantin, X. Zhou, K. Burke, Physical Review Letters, 100 (2008) 136406. [49] S. Goedecker, M. Teter, J. Hutter, Physical Review B, 54 (1996) 1703-1710. [50] M. Bajdich, M. García-Mota, A. Vojvodic, J.K. Nørskov, A.T. Bell, Journal of the American Chemical Society, 135 (2013) 13521-13530. [51] S. Grimme, J. Antony, S. Ehrlich, H. Krieg, The Journal of Chemical Physics, 132 (2010) 154104. [52] J.K. Nørskov, J. Rossmeisl, A. Logadottir, L. Lindqvist, J.R. Kitchin, T. Bligaard, H. Jónsson, The Journal of Physical Chemistry B, 108 (2004) 17886-17892. [53] M. García-Mota, M. Bajdich, V. Viswanathan, A. Vojvodic, A.T. Bell, J.K. Nørskov, The Journal of Physical Chemistry C, 116 (2012) 21077-21082.
Acknowledgements This work was financially supported by the Science Challenge Project (TZ2018004) and National Natural Science Foundation of China (51572016 and U1530401), the National Program for Thousand Young Talents of China. The computational supports from Tianhe-JK at the Beijing Computational Science Research Center (CSRC) and NSFC-Guangdong Joint Fund (the second phase) under Grant No. U1501501 are greatly acknowledged.
Pan Guo received his B. S. degree in Applied chemistry from Lanzhou University, Gansu province, China in 2013. He then joined Beijing Computational Research Center, China Academy Of Engineering
Physics. He is now a Ph. D. candidate in Prof. Li-Min Liu and Woon-Ming Lau’s Research Group. His current research interests include the electro-catalysis and Li-ion battery.
Jian Wu received his Ph.D. degree from Beijing Computational Research Center in 2016. Currently, he is working in School of Mathematics and Physics, University of Science and Technology Beijing. His present scientific interests focus on the theoretical and experimental deign of nanomaterials for energy conversion and storage.
Xi-Bo Li received his Ph.D. degree from Beijing Computational Research Center in 2016, under the supervision of Prof. Li-Min Liu. Currently, he works in Department of Physics, Jinan University. He has co-authored more than 10 journal papers. His current research is mainly focused on theoretical deign of novel materials for fuel cell and photocatalysis.
Jun Luo received his B.S. (2001) and Ph.D. (2006) from Tsinghua University, China. Then, he worked as a postdoc in Warwick University and a research fellow in Oxford University, UK. In 2011, he joined Tsinghua University as an associate professor and was granted "National Program for Thousand Young Talents of China" (namely the 1000-plan for young talents). In 2015, he moved to Tianjin University of Technology and is a full professor and the director of Center for Electron Microscopy with an aberration-corrected STEM and other five electron microscopes there. His research focuses on low-dimensional materials and electron microscopy.
W.M. Lau currently leads the Center for Green Innovation at the University of Science and Technology Beijing in China. He graduated from the Chinese University of Hong Kong (BSc) and the University of British Columbia (PhD). Professor Lau has conducted research in surface science and engineering, materials science and device engineering, nanotechnology, green energy and green manufacturing, and zero-carbon urbanization. He has published over 350 scientific articles and disclosed over 20 patents, with most patents being licensed. He holds a National Thousand-Talents Award in China and a Green Chemistry Award in Canada.
Hao Liu is an associate professor in Chengdu Green Energy and Green Manufacturing Technology R&D Center, Chengdu Development Center of Science and Technology of CAEP. He received his Ph.D. in 2010 from department of mechanical and materials engineering of the University of Western Ontario. After a short period postdoctoral research in Western Surface Science, he joined CAEP. His current research interests lie within the design of novel nanomaterials for clean energy, especially for batteries.
Xue-Liang (Andy) Sun is a Canada Research Chair in Development of Nanomaterials for Clean Energy, Fellow of the Royal Society of Canada and Canadian Academy of Engineering and Full Professor at the University of Western Ontario, Canada. Dr. Sun received his Ph.D. in materials chemistry in 1999 from the University of Manchester, UK, which he followed up by working as a postdoctoral fellow at the University of British Columbia, Canada and as a Research Associate at L’Institut National de la Recherche Scientifique (INRS), Canada. His current research interests are focused on advanced materials for electrochemical energy storage and conversion.
Li-Min Liu received his Ph.D. from Institute of Metal Research, Chinese Academy of Sciences in 2006. During his Ph.D. study, he visited Queen's University of Belfast, UK, for a year. Then he worked in Fritz-Haber-Institut (Germany), University College London (UK), and Princeton University. Since 2012, he has taken a tenure-track 26 / 27 position in Beijing Computational Science Research Center. He has co-authored more than 80 journal papers. He was granted "the 1000-plan for young talents" and "the National Science Fund for Excellent Young Scholars". His research interests focus on nanocatalysts, TiO2-based photocatalysis, electrocatalysis and fuel cells.
Highlights
3D binder-free catalyst was prepared by CVD and electrodeposition.
The Co(OH)2@NCNTs@NF possesses open porous structure.
The Co(OH)2@NCNTs@NF exhibits high activity and stability in water-splitting.