Received 14 January 2016; accepted 22 February 2016; published 25 February 2016
1. Introduction
Hydrogen has been considered as a clean and efficient fuel in the transition from the current hydrocarbon economy because of the climate change and the shortage of fossil fuels [1] -[3] . Though the hydrogen evolution reaction (HER) can be effectively facilitated by Pt-group metals, the high cost and scarcity of Pt-group metals make the widespread application of these catalysts difficult. The exploitation of efficient HER catalysts among low cost and abundant compound is therefore desirable [4] . Successful examples include MoS2 [4] -[7] , WS2 [8] [9] , WS3 [9] , CoS2 [10] , MoSe2 [11] [12] , WSe2 [12] , CoSe2 [13] , MoB [14] , Mo2C [15] , NiMoNx [16] , Co0.6Mo1.4N2 [17] , MoP [18] , Ni2P [19] , Ni12P5 [20] , Co2P [21] , and CoP [22] .
Because most electrode materials suffer from corrosion in acidic condition, alkaline water electrolysis is widely adopted in industry [3] . Nowadays, the Co-based materials have attracted considerable attention due to their high activity toward HER and low cost, especially in the aspect of alkaline water electrolysis. In this study, we show that Co9S8 nanotubes can work as an earth-abundant electrocatalyst with efficient catalytic activity and excellent stability during HER in basic solution.
2. Experimental
2.1. Materials Synthesis
The method used in the synthesis of Co(CO3)0.35Cl0.20(OH)1.10 nanorods was adopted from those reported in ref [23] [24] . Then, 0.6 mmol of the as-prepared Co(CO3)0.35Cl0.20(OH)1.10 and 0.629 mL of a supersaturated Na2S aqueous solution were loaded into a Teflon liner (40 mL) with 30 ml distilled water [24] . The liner was sealed in a stainless steel autoclave and maintained at 160˚C for 8 h, then cooled naturally to room temperature. The black precipitates were filtered off, washed with distilled water and ethanol, and then dried at 60˚C.
2.2. Material Characterization
The morphologies were accessed by scanning electron microscopy (SEM, 7001F, JEOL) and transmission electron microscopy (TEM, 2100, JEOL). The X-ray photoelectron spectroscopy (XPS) experiments were carried out on a Physical Electronics PHI 5700 ESCA System. Powder X-ray diffraction (XRD) patterns were collected with a D8 ADVANCE.
The electrochemical measurements were carried out in an 1M KOH aqueous solution with an electrochemical workstation (CHI614D, CH Instrument). A three-electrode configuration was adopted in the measurements, with Co9S8 loading on GCE as the working electrode, a graphite rodas the counter electrode and a Mercury/Mercury Oxide electrode (MOE, Hg/HgO) as the reference electrode. The reversible hydrogen evolution potential (RHE) was determined to be −0.879 V vs MOE by the open circuit potential of a clean Pt electrode in the same solution. For the evaluation of HER catalytic activity, Co9S8 nanotubes (4 mg) were dispersed in 1 mL of water/ethanol (4/1, V/V) containing 80 μL of Nafion solution (5 wt%). The evaluation of the HER catalytic activity of Co9S8 loaded on GCE was carried out by linear sweep voltammetry (5 mV∙s−1). The volume of H2 during potentiostatic electrolysis measurement was monitored by volume displacement in a configuration as shown in ref [21] .
3. Results and Discussions
The morphology of Co9S8 was examined by SEM. Figure 1(a) shows that nanotubes are arranged in a radial fashion. A SEM image with higher magnification (Figure 1(b)) shows that the surfaces of the nanotubes are very rough, which indicates that they are composed of many tiny nanoparticles.
The overall structural features of Co9S8 was accessed via XRD experiments (Figure 2(a)). The patterns of product are well associated with those of cubic phase Co9S8 (JCPDS No. 65-1765). The hollow structure of nanotubes can be confirmed by TEM image (Figure 2(b)). High resolution (HRTEM) image (Figure 2(c)) shows that the nanotube is composed of tiny nanoparticles with diameters of only several nanometers. The observed interplanar spacing is 0.24 nm and 0.28 nm, which corresponds to the separation between (400) and (222) plane of cubic phase Co9S8, respectively. The selected area electron diffraction (SAED) pattern of nanotubes (Figure 2(d)) shows distinct diffraction rings, which can be indexed as (311), (511), (440), (533), (642) and (931) lattice planes of cubic phase Co9S8.
Chemical states of Co and S were obtained from XPS characterization (Figure 3). Figure 3(a) shows the core level spectrum of the Co 2p region, with Co 2p3/2 binding energies at 778.9, 781.2 and 786.5 eV. The peak at 778.9 eV suggests that there are reduced Co species in Co9S8 [25] . These reduced Co species are partially charged (Coδ+, 0 < δ < 2), and δ must have a small value, because the corresponding Co 2p3/2 binding energy (778.9 eV) is very close to that of metallic Co (777.9 eV) [21] [26] . In the S 2p spectrum (Figure 3(b)), the peak centered at 161.9 eV agrees with the binding energies of Co?S [27] [28] .
Figure 4(a) shows the representative polarization data of Co9S8/GCE electrodes with different mass loadings, along with the polarization data of a bare GCE and commercial Pt/C catalyst (Johnson Matthey, Hispec 3000, 20 wt%) loaded on GCE. Co9S8/GCE electrodes with different loading amounts of Co9S8 nanotubes all show apparent current density in the potential range of 0 to −0.4 V vs RHE. The electrocatalytic activity of the Co9S8
Figure 1. (a) Low and (b) high magnification SEM images of the Co9S8 nanotubes.
Figure 2. (a) XRD patterns (b) TEM image (c) HRTEM image, and (d) SAED pattern of the Co9S8 nanotubes.
Figure 3. XPS spectra of (a) the Co 2p3/2 and (b) the S 2p windows of Co9S8 sample.
sample is more efficient when the loading amount increased in some degree. The sample with optimal performance (loading amount: 0.855 mg∙cm−2) has a current density of 20 mA∙cm−2 at an overpotential of 320 mV. However, when the loading amount is further increased, the Co9S8 doesn’t result a better electronic property for the electrode because it may result in large interface resistance. In contrast, negligible current can be found from the bare GCE electrode, showing that the large current in Co9S8/GCE can be definitely correlated with Co9S8 nanotubes. The performance of representative HER catalysts is summarized in Table S1 (Electronic Supplementary Information), which shows that the performance of Co9S8 nanotubes is superior to that of Ni3S2 loaded on carbon nanotubes and Fe2P/NGr nanocomposite loaded on GCE.
In a potentiostatic electrolysis experiment, the time-dependent current density recorded from a potentiostatic electrolysis shows only a little degradation in 20,000 s (Figure 4(b)). In an accelerated degradation experiment, cyclic voltammetry (CV) sweeps were carried out in the 1 M KOH aqueous solution between −0.380 and 0.100 V versus RHE (inset of Figure 4(b)). It shows that after continuous cyclic voltammetric (CV) scan for 4000 cycles, the overpotential required for the current density of 20 mA∙cm−2 (η20) increases from 330 mV to 350 mV. These results suggest Co9S8 nanotube can afford long-term hydrogen generation in basic condition.
The faradaic efficiency of Co9S8 nanotube in electrolysis of water was evaluated by water-displacement method. Figure 4(c) shows the comparison of the theoretical volume of hydrogen and the experimentally measured volume of hydrogen. It is shown that the faradaic yield of H2 production in a potentiostatic electrolysis of water using Co9S8 nanotube as a HER catalyst is nearly 100%.
The Nyquist plots of the Co9S8 nanotubes at overpotentials from −200 to −370 mV (Figure 4(d)) exhibit classic two time-constant behavior. The semicircles at high frequencies can be related to the contact between the catalyst (Co9S8) and the GCE, while those at low frequencies are correlated to the kinetics of the HER process on the surface of the catalyst. The kinetics of electrochemical reaction at an electrode’s surface is usually assessed by charge transfer resistance (Rct), with a smaller Rct value corresponding to faster kinetics. Rct was be deduced from EIS spectra by data fitting, in the present case using the equivalent circuit shown in the inset of Figure 4(d). In Figure 4(e), the applied potential is plotted versus the inverse Rct on a logarithmic scale, and a Tafel slope was determined to be 135 mV∙dec−1 according to the slope of linear portion in the plot. The Tafel slope of 135 mV∙dec−1 suggests that a Volmer-Tafel mechanism is responsible for the HER process on the surface of Co9S8 nanotubes.
Figure 4. (a) Polarization curves of Co9S8 nanotubes and commercial Pt/C catalyst loading on GCE. (b) The time-dependent current density of Co9S8 under overpotential of 280 mV for 20,000 s. The inset shows Polarization curves of Co9S8 nanotube corresponding to the initial and 4000th CV scans. (c) Current efficiency for H2 production using Co9S8/GCE. (d) Nyquist plots of the Co9S8 nanotubes recorded at different overpotentials in 1M KOH. The inset shows the equivalent circuit used for data fitting. (e) Semi-logarithmic plot of applied potential vs log(Rct−1). Only potentials in (a) were corrected with iR drop.
4. Conclusion
In summary, Co9S8 nanotubes are found to be an effective HER electrocatalyst. The optimal η20 is as small as 320 mV in basic solution. Co9S8 nanotubes can work stably in alkaline solutions, and the faradic yield during electrolysis is nearly 100%. The HER process follows a Volmer-Heyrovsky mechanism. The results presented here further demonstrate the promising application potential of metal sulfide in the field of hydrogen generation from electrolysis of water.
Acknowledgements
This research was financially supported by the National Natural Science Foundation of China (61006049, 50925207), the Ministry of Science and Technology of China (2011DFG52970), the Ministry of Education of China (IRT1064), 111 Project (B13025), Jiangsu Innovation Research Team, Jiangsu Province (2011-XCL-019 and 2013-479), and Natural Science Foundation of Jiangsu (BK20131252).
Supplement: Electronic Supporting Information
S1. Experimental
S1.1. Materials Synthesis
The method used in the synthesis of Co(CO3)0.35Cl0.20(OH)1.10 nanorods was adopted from those reported in ref [1] [2] . Then, 0.6 mmol of the as-prepared Co(CO3)0.35Cl0.20(OH)1.10 and 0.629 mL of a supersaturated Na2S aqueous solution were loaded into a Teflon liner (40 mL) with 30 ml distilled water [2] . The liner was sealed in a stainless steel autoclave and maintained at 160˚C for 8 h, then cooled naturally to room temperature. The black precipitates were filtered off, washed with distilled water and ethanol, and then dried at 60˚C.
S1.2. Material Characterization
The morphologies were accessed by scanning electron microscopy (SEM, 7001F, JEOL) and transmission electron microscopy (TEM, 2100, JEOL). The X-ray photoelectron spectroscopy (XPS) experiments were carried out on a Physical Electronics PHI 5700 ESCA System. Powder X-ray diffraction (XRD) patterns were collected with a D8 ADVANCE.
The electrochemical measurements were carried out in an aqueous 1M KOH solution with an electrochemical workstation (CHI614D, CH Instrument). A three-electrode configuration was adopted in the measurements, with Co9S8 loading on GCE as the working electrode, a graphite rodas the counter electrode and a Mercury/Mercury Oxide electrode (MOE, Hg/HgO) as the reference electrode. The reversible hydrogen evolution potential (RHE) was determined to be −0.879 V vs MOE by the open circuit potential of a clean Pt electrode in the same solution. For the evaluation of HER catalytic activity, Co9S8 nanotubes (4 mg) were dispersed in 1 mL of water/ethanol (4/1, V/V) containing 80 μL of Nafion solution (5 wt%). The evaluation of the HER catalytic activity of Co9S8 loaded on GCE was carried out by linear sweep voltammetry (5 mV∙s−1). The volume of H2 during potentiostatic electrolysis measurement was monitored by volume displacement in a configuration as shown in ref [3] .
Table S1. Summary of the HER performance of representative catalysts.
NOTES
*Corresponding author.