Physicochemical characterization
The morphology of the as-prepared samples was characterized by scanning electron microscopy. As shown in Fig. 1a, on the surface of sample TiO2/CoMoO4 the TiO2 nanotubes are not visible (pure TiO2 nanotubes are presented in Figure S1) because during the hydrothermal process the CoMoO4 particles densely covered TiO2 surface.
Intriguingly, each CoMoO4 microspheres (Fig. 1b) is actually a three dimensionally (3D) interconnected porous structure and is assembled from numerous nanospheres. For TiO2/CoMoO4 sample the diameter of CoMoO4 particles ranges from 1 to 2 µm. It can be observed that the particles agglomerate. The surface of TiO2/SrMoO4 sample (Fig. 1c) clearly demonstrated that the product considered of a large amount of spherical structures covered TiO2 surface and shows the similar morphology to TiO2/CoMoO4 sample. The diameters of SrMoO4 spheres are about 2 µm. A higher-magnification SEM image (Fig. 1d) reveals that an individual sphere is composed of tens of similar nanosheets. These nanosheets are connected with each other to form a sphere with random orientation. Figures 1e and 1f illustrate the SEM images for TiO2/SrMoO4/CoMoO4 sample with different magnifications. Figure 1f shows the high magnification from which it can be seen that particles deposited on the TiO2 surface do not possess a uniform size and shape. The mean diameters of 3D microspheres are greater and fluctuate between 2−4 µm. These nanosheets are connected with each other to form nanosheets-based microstructures with random orientation. Furthermore, the surface of some of the crystals is very smooth. In addition, the energy dispersive X-ray analysis also confirms the SrMoO4 and CoMoO4 structure (Figure S2).
The morphologies and microstructures of the hybrid materials include of TiO2, SrMoO4 CoMoO4, and GO are present in Fig. 2. These images confirm that applying hydrothermal technology can prepare in one sample three-dimensional (3D) structures with different morphology. Figure 2a for TiO2/SrMoO4/GO sample indicates the SrMoO4 are randomly distributed on the TiO2 surface. Figure 2b displays that the as-prepared sample TiO2/CoMoO4 has microstructure with CoMoO4 agglomerated on the surface. It can also be found that these microspheres are consisted of a large number of nanosheets.
The surface morphology of CoMoO4 with GO is comparable to sample without GO. The agglomeration of particles impacts on their size distribution and the average diameter is 1 µm. For sample TiO2/SrMoO4/CoMoO4/GO (Fig. 2c) addition of GO caused to stronger agglomeration of SrMoO4 and CoMoO4 to spherical structures. To visualize the distribution of Co, Mo, Sr, Ti, O, S, and C, energy dispersive X-ray spectroscopy mapping was performed (Figure S3).
The obtained materials were also structurally characterized by XRD. Figures 3 and 4 shows the XRD pattern of TiO2/SrMoO4, TiO2/CoMoO4, TiO2/SrMoO4/CoMoO4, TiO2/SrMoO4/GO, TiO2/CoMoO4/GO, TiO2/SrMoO4/CoMoO4/GO. Some of the SrMoO4 and CoMoO4 peaks overlap with the intense anatase peaks. However, several characteristic peaks from these compounds can be observed, as indicated by the corresponding symbols in Figs. 3 and 4. However, the low intensity of these peaks proves the low crystallinity of SrMoO4 and CoMoO4. Nonetheless, the signature peaks of SrMoO4 at 33.7, 37.1, 38.5, 48.1, 76.3, 77.4, 82.3, 86.824–26 and of CoMoO4 at 25.4, 32.4, 36.3, 38.5, 40.2, 45.2, 63.027–29 are very prominent in all samples30,31. The remaining crystallite phases were indexed as characteristic peaks of anatase and the titanium (Ti) phase acting as TiO2 NT support32.
The ornamental GO structure peak was observed at 10.5 plane for only a TiO2/SrMoO4/GO hybrid. However, the presence of GO can be better characterized by Raman spectroscopy.
Raman spectra of TiO2 nanotubes, TiO2/SrMoO4/CoMoO4, and TiO2/SrMoO4/CoMoO4/GO within the frequency range 100–3300 cm−1 are shown in Fig. 5. Several bands characteristic for the pure anatase crystalline form were identified in samples of pristine TiO2 and the hybrid materials. However, in the case of a material with added carbon, the anatase peaks are not that pronounced. The bands located at 143, 398, 516, and 640 cm− 1 are attributed to Eg(1), B1g, A1g, and Eg(3) active anatase modes, respectively33,34. The band at around 326 cm− 1 is attributed to the symmetric stretching of the Co–O–Mo bond35. The peak of around to 300 cm− 1 corresponds to Sr–O–Mo bond36. The band located at 802 cm− 1 is associated with asymmetric stretching modes of O–Mo–O bond while the band at 904 cm− 1 corresponds to the symmetric stretching mode of Mo–O bond37. On analyzing the Raman bands of modified material by GO, two distinct bands, namely, D and G bands, were obtained at 1347 cm− 1 and 1580 cm− 1.
The D band is corresponding to disorder carbon while the G band is attributed to sp2 hybridized carbon38,39. This confirmed that the graphene component is maintained during the hydrothermal process. Moreover, the second order of zone boundary phonons or 2D band which is related to the stacking nature of graphene layers was observed at 2706 cm− 1 for GO40.
Electrocatalytic Activities For Hydrogen Evolution
The catalytic activity of obtained hybrids for hydrogen evolution reaction (HER) was measured in 0.2 M H2SO4 using three-electrode configuration with a scan rate of 5 mV s−1. For comparison, bare TiO2 nanotubes and Pt were also tested. Figures 6a and 6b give the HER polarization curves of commercial Pt disc, pure TiO2 nanotubes, and the obtained hybrids without and with the addition of carbon.
According to linear sweep voltammograms (LSVs) in Figs. 6a and 6b, the overpotential values for all electrodes at jHER=10 mA cm−2 are summarized in Table 1. The obtained values indicate that a small amount of carbon has a large share in the HER performance of the obtained hybrids. It is worth noting that the η value of TiO2/SrMoO4/CoMoO4/GO is only 29 mV larger than that of pure Pt and comparable to or smaller than those of various transition metal-based electrocatalyst (Table S1).
Table 1
The HER parameters and Tafel slopes for the obtained catalysts.
Electrode | Overpotential (mV vs. RHE) to achieve a current density of -10 mA cm− 2 | Onset potential (mV vs. RHE) | b (mV dec− 1) |
Pt | 92 | -75 | 38 |
TiO2 | does not achieve | - | 330 |
TiO2/SrMoO4 | 313 | -275 | 152 |
TiO2/CoMoO4 | 242 | -219 | 168 |
TiO2/SrMoO4/CoMoO4 | 167 | -209 | 91 |
TiO2/SrMoO4/GO | 146 | -96 | 62 |
TiO2/CoMoO4/GO | 184 | -121 | 75 |
TiO2/SrMoO4/CoMoO4/GO | 121 | -60 | 90 |
Figure 7 presents the Tafel plots. At high overpotentials the HER on the electrodes is kinetically controlled, which can be given by the Tafel Eq. 41:
where η (V) means overpotential, a (V) is the cathodic intercept related to the exchange current density, b (V dec− 1) means the cathodic Tafel slope, and j (A cm− 2) means catalytic current density. Tafel slope values calculated from the linear portion of potential vs. logarithmic value of current density deliver useful kinetic metrics of the catalyst30. The Tafel slope values were 38, 330, 152, 168, 91, 62, 75, and 90 for Pt, TiO2 nanotubes, TiO2/SrMoO4, TiO2/CoMoO4, TiO2/SrMoO4/CoMoO4, TiO2/SrMoO4/GO, TiO2/CoMoO4/GO, TiO2/SrMoO4/CoMoO4/GO, respectively. Concerning as-synthesized electrocatalysts, TiO2/SrMoO4/GO showed the lowest Tafel slope value. In this case, the positive effect of the presence of carbon in all hybrids was also observed. The data of Tafel plots along are also tabulated in Table 1.
To examine the long-term stability of obtained electrodes, chronopotentiometric tests were also performed at the fixed current density of 10 mA cm−2 (Figs. 8a and 8b). As can be seen, almost all electrodes show good stability of overpotential with a polarization current of -10 mA cm−2. However, in the case of pure TiO2 nanotubes the overpotential had shifted from 842 mV to 541 mV by the end of 200 min. In other cases, the overpotential oscillates between 261 and 140 mV.