Yan
Liu‡
a,
Jiao
Li‡
b,
Dan
Sun
a,
Linglan
Men
a,
Bo
Sun
a,
Xiao
Li
*a,
Qingbo
An
*a,
Fangbin
Liu
a and
Zhongmin
Su
*a
aSchool of Chemical and Environmental Engineering, Jilin Provincial Science and Technology Innovation Centre of Optical Materials and Chemistry, Changchun University of Science and Technology, International Joint Research Center for optical functional materials and chemistry, Changchun University of Science and Technology, Changchun, People's Republic of China. E-mail: lix@cust.edu.cn; zmsu@nenu.edu.cn
bSchool of Materials science and Engineering, Changchun University of Science and Technology, Changchun 130022, People's Republic of China
First published on 19th November 2021
The design of economical and efficient electrocatalysts to replace precious platinum-based catalysts for hydrogen evolution reaction (HER) has always been an essential study direction of clean energy technology. In this work, an HER electrocatalyst (Co/WC@NC) was synthesized by pre-self assembly of polyoxometalate (Co4P4W30), cobalt ions and dicyandiamide (DCA) via a one-step calcination process. Co/WC@NC consists of cobalt and tungsten carbide nanoparticles, which are wrapped with nitrogen-doped carbon layers. Co/WC@NC shows good HER capability with relatively low overpotentials of 142 and 158 mV with small Tafel slopes of 93 and 95 mV dec−1, respectively, at a current density of 10 mA cm−2 in acidic and alkaline solutions. Moreover, Co/WC@NC provides satisfactory durability for 24 h in both 0.5 M H2SO4 and 1.0 M KOH. Founded on the above results, good catalytic property of Co/WC@NC can be primarily associated with the synergy between Co and WC nanoparticles, N element doping to make adjustments for the electronic structure of carbon layers, thus speeding up the electron transfer process, as well as graphite-coated carbon layers protecting nanoparticles from oxidation, and improving stability. This study confirms the universal idea to form transition carbides for an efficient electrocatalytic production of hydrogen.
Currently, non-precious electrocatalysts are mainly focused on carbides,16,17 phosphides,18–20 nitrides,21–23 sulphides,24 and a series of N-type non-metallic carbon materials.25 Thereinto, transition metal carbides (TMCs) have been widely investigated due to the special electronic configuration as well as good electrical conductivity, and similar chemical activity and catalytic performance with platinum-based catalysts in HER fields, especially in the VB/VIB family.26,27 Moreover, compared with other transition metals, the catalytic activity of W-based carbides for HER needs to be further improved.28 Recently, a few numbers of tungsten-based carbides have been reported. For instance, Zhang and co-works report dual-phased Mo2C–WC ultrathin nanosheet nanocrystals (Mo2C–WC/NCAs) through one-step pyrolysis of polymers.29 Yan and co-workers synthesized ultrafine WC by in-suit molten salt.30
Polyoxometalates (POMs), as one class of plentiful metal oxygen clusters, are composed of transition metal ions (Mo, W, V, Nb, Ta, etc.).31 POMs possess large number of negative charges with peculiar chemical compositions and structural diversification, which can be regarded as one of the ideal candidates for synthesizing uniform Mo/W-based electrocatalyst.32–35 Tungsten-based polyoxometalates are widely studied because of their unique nanostructures and inherent high conductivity,30 especially in preparing tungsten carbides. Whereas the high intensity of W–H can seriously affect H2 desorption, leading to poor HER activity. According to previous studies, the strength of the W–H bond can be reduced by adjusting the electronic structure of WC, such as by introducing a second metal source.36 Furthermore, the concurrent doping of N, S, and P heteroatoms also help to regulate the electronic structure. The abundant N elements in dicyandiamide (DCA) molecules show high electronegativity, which can accomplish the doping of N atoms to change the local electron configuration of carbon materials, promoting charge transfer, and accordingly improve the catalytic capability. Therefore, dicyandiamide (DCA) can be used as a potential carbon and nitrogen source to synthesize catalysts for electrochemical applications.37
In this work, Co2+ and POM (Co4P4W30) are selected as metal sources and DCA as nitrogen sources and carbon sources for the synthesis of hydrogen evolution catalysts (Co/WC@NC). Co/WC@NC consists of nitrogen-doped carbon-coated Co and WC nanoparticles. Thereby, under acidic conditions, Co/WC@NC only needs 142 mV to attain 10 mA cm−2, fulling of the Tafel curvature of 93 mV dec−1. Besides, the overpotential and Tafel slope of Co/WC@NC are 158 mV and 95 mV dec−1 in basic solutions. Graphitic carbon coating enhances the steadiness of the catalyst in either acid or alkaline solutions so that Co/WC@NC presents satisfactory durability for 24 h in both 0.5 M H2SO4 and 1.0 M KOH. Co and WC nanoparticles synergistically enhance charge transfer rate and improve the HER performance. Moreover, the successful doping of nitrogen will further enhance the catalytic activity Co/WC@NC. The work indicates that the development of tungsten-based materials to study electrocatalytic hydrogen evolution exhibits large research spaces for potential practical applications.
Fig. 1 shows the transmission electron microscope (TEM) and energy spectrum (EDS) mapping images of Co/WC@NC. These images exhibit the lattice structure, configuration of surface, and particle size of the catalyst. Co/WC@NC showcases an irregular modality with particle sizes ranging from 10 nm to 40 nm (Fig. 1a). The higher resolution transmission electron microscope (HRTEM) image (Fig. 1b) indicates that Co/WC nanoparticles are wrapped by graphitic carbon layers, and the crystal grating fringe spacing of Co and WC are 0.20 nm and 0.26 nm, respectively; correspondence with the lattice plane (111) in Co (JCPDS No. 15-0806) and the grate plane (100) in WC (JCPDS No. 51-0939). The element mapping graphs of Co/WC@NC exhibits that the elements (Co, W, C, N and P) are well-dispersed in the catalyst (Fig. 1c–h). The mass percentages of C, N, P, Co and W in EDS are 90.61, 5.23, 1.33, 1.97 and 0.87%, respectively. In addition, the atomic percentages of Co, W, C, N and P (Co/WC@NC, WC@NC and Co@NC) from XPS are shown in Table S1 (ESI†). The ICP data indicate the mass percentages of C, N, P, Co and W in Co/WC@NC as 93.152%, 4.358%, 0.350%, 0.689% and 1.451%, respectively. The presence of nitrogen attests that nitrogen is thrivingly deathniumed into Co/WC@NC.
![]() | ||
Fig. 1 Characterization Morphology of Co/WC@NC: (a) TEM image of Co/WC@NC; (b) HRTEM image; (c–h) EDS mapping images. |
To characterize and analyze the phase composition of Co/WC@NC, powder X-ray diffraction (PXRD) patterns were recorded. It can be seen from Fig. 2 that Co/WC@NC reveals two phases: Co and WC. The main diffraction peaks of WC (JCPDS No. 51-0939) are located at 31.5°, 35.6°, 48.3°, 64.0°, 65.8°, 73.1°, 75.5° and 77.1°, corresponding to the lattice planes (001), (100), (101), (110), (002), (111), (200) and (102), respectively. The obvious peaks at 44.1°, 51.4° and 75.8° correspond to the diffraction peaks of crystal planes (111), (200) and (220) in Co (JCPDS No. 15-0806). It is highlighted that the wider peak with the scope of 20°–30° (2θ) was not evident, maybe owing to the density of the porous carbon produced by dicyandiamide being relatively low.39 Simultaneously, we also considered PXRD patterns of the comparative samples (Fig. S1, ESI†). WC/W2N@NC shows two phases, WC and W2N. The clear diffraction peaks of W2N (JCPDS No. 25-1257) are located at 37.7°, 43.8°, 63.7° and 76.5°, which correspond to (111), (200), (220) and (311) planes, respectively. Co@NC possesses the same peaks as Co/WC@NC. In addition, Raman spectra were used to investigate the degree of graphitization Co/WC@NC (Fig. S2, ESI†). The two peaks at 1326 and 1592 cm−1 represent the d-band and g-band of the disordered and graphitic carbon in Co/WC@NC, respectively. The peak-to-strength ratio of ID/IG was 1.195, indicating that the structure demonstrates defects and the degree of graphitization is equivalent, which is conducive to the absorption of H+/H2.40
X-Ray photoelectron spectroscopy (XPS) measurements were performed on Co/WC@NC to analyze the element composition and valence state. Fig. 3a reveals the XPS spectrum of Co 2p in the Co/WC@NC catalyst. The deconvolution spectrum of Co 2p defines the existence of Co2+ 2p3/2 (780.5 eV), Co2+ 2p1/2 (796.0 eV), metallic Co (778.0 and 793.0 eV), and satellite peaks (786.0 and 802.9 eV).41,42 The peak area of cobalt metal is higher than that of other peaks. The more valence state of Co is beneficial to encourage electron shift and convenient to combine with W, and arouses the catalytic efficiency of activator. In W 4f deconvolution spectrum of Co/WC@NC (Fig. 3b), strong diffraction peaks of W 4f7/2 (32.2 eV) and W 4f5/2 (34.4 eV) may be attributed to the charged-off WC.43 Moreover, the other W 4f7/2 (35.2 eV) and W 4f5/2 (37.3 eV) peaks are all subordinated to the partial oxygenation of W. The XPS spectrum of C 1s (Fig. S3a, ESI†) can be deconvoluted into three peaks (288.8 eV, 286.5 eV and 284.6 eV), belonging to CO, C–N and C
C, respectively.44 As can be seen from XPS spectra of N 1s (Fig. S3b, ESI†), the peaks at 401.5, 400.5 and 398.4 eV are in agreement with graphite N, pyrrole N and pyridine N, respectively.45 These results indicate that the N element was added into the lattice of C layers. Fig. S4 and S5 (ESI†) show the XPS spectrum of WC@NC and Co@NC, respectively. The full XPS spectrum of W 4f in the WC@NC verifies the existence of W 4f7/2 (31.7 and 36.5 eV), W 4f5/2 (34.7 and 41.0 eV) and satellite peaks (44.0 eV) (Fig. S4a, ESI†).46 The Co 2p spectrum (Fig. S4b, ESI†) of WC@NC demonstrates six peaks at 778.0, 780.2, 785.4, 793.1, 796.0 and 803.0 eV. Among them, the peaks at 785.4 and 803.0 eV are satellite peaks attributed to the vibration excitation of Co2+ ions. The peaks at 778.0, 780.2 eV and 793.1, 796.0 eV can result from tCo 2p3/2 and Co 2p1/2 of metallic Co and cobalt ions.47,48
Fig. S4c (ESI†) reports the peaks corresponding to CC (284.5 eV), C–N (286.4 eV) and C
O (288.6 eV) of C 1s. The peaks at 397.8, 399.6 and 401.3 eV in shown in Fig. S4d (ESI†) are due to graphitic-N, pyrrolic-N and pyridinic-N of N 1s. Fig. S5a, b and c (ESI†) show Co, N and C XPS spectra of Co@NC. The metallic Co (778.0/793.1 eV), Co2+ 2p3/2 (780.7 eV), Co2+ 2p1/2 (796.2 eV) and satellite peaks (786.7/802.5 eV) are clearly seen in the Fig. S5a (ESI†).49 C 1s exists in C
C (284.9 eV), C–N (286.0 eV) and C
O (288.6 eV) bond. The peaks at 401.8, 400.1 and 397.7 eV of N 1s can be assigned to graphitic-N, pyrrolic-N and pyridinic-N. To evaluate the porosity of Co/WC@NC, N2 adsorption–desorption measurement was conducted. As shown in Fig. S6 (ESI†), the isotherm obtained from N2 adsorption exposes an obvious hysteresis loop, indicating that the catalyst belonging to the H3 type. The BET surface area of Co/WC@NC is 72.2 m2 g−1 and the pore size is 3.97 nm, presenting mesoporous structures. The mesoporous structure facilitates more active site exposure, thereby enhancing the electrocatalytic activity.
In 0.5 M H2SO4 and 1.0 M KOH, a universal three-electrode system was used to preliminarily explore the HER catalytic property of Co/WC@NC at room temperature. The electrocatalysts were uniformly coated on the surface of glassy carbon electrodes with a capacity of 0.66 mg cm−2, and under the condition of a particular scan rate (5 mV s−1). After the linear sweep voltammetry (LSV) test, the polarization curves were obtained. For comparison, different calcination temperatures, distinct ratios and mass loadings were tested. Above all, we selected a proportion of Co4P4W30/DCA corresponding to the mass ratio of 1:
7. The LSV curves at different temperatures such as 700, 800 and 900 °C under the condition of (Co4P4W30/DCA) mass ratio of 1
:
7 were gauged (Fig. S7, ESI†). Simultaneously, XRD patterns of the catalyst with different temperatures (700, 800 and 900 °C) based on the ratio of Co4P4W30 to DCA as 1
:
7 are shown in Fig. S8 (ESI†). The diffraction peak at 35.6° and 48.3° in samples heated at 700 and 800 °C are ascribed to (100) and (101) crystal planes of WC (JCPDS No. 51-0939). The sample heated at 900 °C exists two phases, WC (JCPDS No. 51-0939) and W2C (JCPDS No. 35-0776). By comparing different calcination temperatures, the catalyst at 700 °C showed better performance. In order to optimize the experiments, we then prepared electrocatalysts with different Co4P4W30/DCA mass ratios as 1
:
4, 1
:
5, 1
:
6, 1
:
7 and 1
:
8 at the optimised temperature and tested the LSV curves (Fig. S9, ESI†). The best quality ratio was 1
:
7, but it still does not achieve the expectation. The ratios of Co4P4W30 to Co(NO3)2 (1
:
1, 1
:
2, 1
:
3, 3
:
1 and 2
:
1) were conducted under the condition of 700 °C and the mass ratio (Co4P4W30/DCA) of 1
:
7. Through the polarization curve test, the optimal mass ratio was 1
:
2, as shown in Fig. S10 (ESI†). It only needed an overpotential of 142 mV in acidic solution to reach a current density of 10 mA cm−2. After a series of sample comparison, only when the mass ratio of Co4P4W30/DCA/Co(NO3)2 was 1
:
7
:
2, the electrocatalyst demonstrated the best electrochemical activity. Besides, loading adjustment is the last step of the catalyst condition optimization. The LSV curves of different mass loadings are shown in Fig. S11 (ESI†). 0.66 mg cm−2 was the optimal loading of the catalyst.
For further analyzing the HER property of Co/WC@NC, Co@NC and WC@NC were also prepared. Fig. 4a, Pt/C shows the lowest overpotential of 26 mV (10 mA cm−2) in 0.5 M H2SO4, which is consistent with previous literature reports.20Co/WC@NC exhibits a relatively good catalytic capability, requiring an overpotential of 142 mV when the current density was 10 mA cm−2. Furthermore, WC@NC and Co@NC required 233 and 273 mV, respectively, both of which exhibited under par HER performance than Co/WC@NC. The above results show the mutualism between Co and WC increases the catalytic function of Co/WC@NC. The catalysts mechanism was analysed by calculating Tafel slope values as shown in Fig. 4b. The results are 93 (Co/WC@NC) and 25 mV dec−1 (Pt/C), suggesting the Volmer–Heyrovsky mechanism, which is equivalent to the desorption as the decisive step for hydrogen production.50,51 The Tafel slopes of the comparative catalysts WC@NC and Co@NC are 99 and 135 mV dec−1, respectively.
As we know, the electrochemical activity is associated with the electrochemical superficial area (ECSA) of the electrocatalyst. ECSA is usually calculated from two thicknesses electrochemical capacitance values (Cdl) using a simple cyclic voltammetry (CV) test. Here, the different scan rates of CV curves were 5, 10, 25, 50 and 100 mV s−1 in 0.5 M H2SO4 and 1.0 M KOH conditions, respectively (Fig. 4c, Fig. S12a and c, ESI†). Fig. S12a and c (ESI†) clarify the CV curves of WC@NC and Co@NC, respectively, in 0.5 M H2SO4 solution, and WC@NC is stronger than Co@NC. The values of Cdl were derived from Δj (ja–jc) at different scan rates at 0.1 V (vs. RHE), and the curve was fitted from the obtained points. The Cdl of the specimen was decided from the slope of the curve fitting. The Cdl value of Co/WC@NC under acidic conditions was 9.8 mF cm−2, more than that of Co@NC (4.1 mF cm−2) and WC@NC (5.3 mF cm−2), meaning that Co/WC@NC possesses a larger surface area and exposes numerous electrochemically-capabile sites, as shown in Fig. 4d.
Besides, electrochemical impedance spectroscopy (EIS) was used to assess the electrical conductivity, reflected by the accuse transfer resistance (Rct) value. Fig. 4e shows the Nyquist plot of the EIS of Co/WC@NC and the comparative samples, with an overpotential of 150 mV. The semicircle of Co/WC@NC is the smallest, displaying that Co/WC@NC exhibits the best electrical conductivity. Equally importantly, permanence is an essential basis for judging the suitability of electrocatalysts. The initial LSV curves before and after 2000 cycles are substantially not changing much, as shown in Fig. 4f. Furthermore, the inset of Fig. 4f is the durability test diagram of Co/WC@NC in acidic solution at 150 mV overpotential for 24 h. These results indicate that Co/WC@NC show good electrocatalytic stability.
In addition, we also tested the hydrogen evolution performance of all specimens in alkaline solutions. Undoubtedly, Pt/C was still the best catalyst for HER. The overpotential of Co/WC@NC is relatively low competed with Co@NC and WC@NC. Thereinto, the overpotentials (Tafel slopes) of Co@NC, WC@NC and Co/WC@NC are 216 mV (127 mV dec−1), 175 mV (133 mV dec−1) and 158 mV (95 mV dec−1), respectively, when the current density was 10 mA cm−2 (Fig. 5a and b). The CV curves are illustrated in Fig. 5c, and Fig. S12b and d (ESI†). The Cdl values of Co@NC, WC@NC and Co/WC@NC are 3.7, 4.3 and 5.4 mF cm−2, respectively, displaying that Co/WC@NC manifests a larger electrochemical surface area in alkaline environment (Fig. 5d). EIS spectra under the alkaline conditions with the overpotential of 0.2 V shows that the impedance worth of Co/WC@NC is smaller than that of other comparison samples, indicating that Co/WC@NC exhibits higher conductivity as shown in Fig. 5e. Fig. 5f shows that the catalyst presents good stability in alkaline media. In addition, the antithesis of HER catalysts of bimetallic carbides in recent years is encapsulated in Table S2 (ESI†). The excellent HER performance of Co/WC@NC is primarily ascribed to (1) the mutualism of Co and W in the catalyst. (2) It should be pointed out that the existence of a carbon layer facilitates porosity and conductivity of the whole catalyst. (3) What matters most is that N doping into the carbon layer raises the electron density of the graphite carbon shell and further promotes HER capability.
Footnotes |
† Electronic supplementary information (ESI) available: Experimental section, characterizations, electrochemical measurements, XRD, IR spectras, Raman spectra, XPS, N2 adsorption–desorption isotherms, LSV, CV, Cdl and EIS. See DOI: 10.1039/d1nj04573c |
‡ Contributed equally. |
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2022 |