Xu-Lei Suiab,
Zhen-Bo Wang*a,
Yun-Fei Xiaa,
Min Yanga,
Lei Zhaoa and
Da-Ming Gu*b
aSchool of Chemical Engineering and Technology, Harbin Institute of Technology, No. 92, West-Da Zhi Street, Harbin, 150001, China. E-mail: wangzhb@hit.edu.cn; Fax: +86-451-86418616; Tel: +86-451-86417853
bSchool of Science, Harbin Institute of Technology, No. 92, West-Da Zhi Street, Harbin, 150001, China
First published on 31st March 2015
In this paper, we report a rapid method to synthesize titania nanotubes as the support for a Pt-based catalyst. The titania nanotubes can be obtained during 1200 s in an ethylene glycol system by the anodization method. Pt nanoparticles were successfully deposited on a mixture of carbon and as-prepared TiO2 nanotubes by a microwave-assisted polyol process. The electrochemical results show that the electrochemically active specific surface area and the activity for methanol electrooxidation of the as-prepared catalyst are both much higher than those of the commercial Pt/C. Whether it is through the constant potential test or cycling potential test, the durability of the as-prepared catalyst is higher than that of the commercial Pt/C. Such remarkable performance is due to the strong corrosion resistance of titania, metal–support interactions and hydrogen spillover effect between Pt and titania, the better electronic conductivity, as well as the good dispersion of the Pt nanoparticles. These studies indicate that titania nanotubes are a promising catalyst support for methanol electrooxidation.
The widely accepted method is to disperse Pt nanoparticles on large surface area supports.6,7 Currently, carbon materials are frequently used for catalyst supports, including carbon black,8–10 carbon nanotubes11,12 and graphene.13–15 However, carbon corrosion leads to a decrease of the catalytic performance during the long term operation of DMFCs, especially in the cases where active platinum is present.16–18 Metal oxide supports are desirable supports due to not only the corrosion resistance in acidic and oxidative environments but the strong metal–support interactions with platinum, such as WO3,19 SnO2,20–22 TiO2,23–25 CeO2,26,27 and so on. Metal–support interactions have been studied in detail by Lewera28 and Zhang.29 Further research performed by Jaksic30 showed the enhancement of the electrocatalytic performance due to the spillover effect for interactive hypo-d-oxide supports. Titania has attracted increasing attention due to its low cost and environmental friendliness.31 Nevertheless, the poor conductivity of titania restricts its application in fuel cells. In our previous work, it was confirmed that titania nanotubes can improve the electron conductivity of titania.32 But the synthesis time of the titania nanotubes was up to 3 h, and this is too long for practical applications. In this work, we rapidly fabricated titania nanotubes during 1200 s in an ethylene glycol system by the anodization method. Thus, the usability of titania nanotubes for fuel cells is greatly improved. The Pt-based catalyst was prepared by a microwave-assisted polyol process and was characterized by physical and electrochemical measurements. These studies have shown that the as-prepared catalyst exhibits better activity and durability for the methanol electrooxidation reaction than that of the commercial Pt/C.
The working electrode was fabricated as follows. 2.0 mg mL−1 catalyst ink was formed by the catalyst and ultrapure water. 5 μL catalyst ink was then transferred by pipette onto the glassy carbon electrode. After drying, 5 μL dilute aqueous Nafion® solution was applied onto the surface of the catalyst layer to protect the catalyst from detaching.
Cyclic voltammograms were recorded from −0.64 to 0.51 V vs. MSE at the rate of 50 mV s−1. Before the measurements, the working electrode was first activated until a steady CV curve was obtained. The electrochemically active surface areas (ESA) of platinum were calculated with the formula ESA = QH/(0.21 × MPt).33 The amperometric i–t curves were obtained at a constant potential of 0.6 V vs. NHE for 3600 s in an acidic methanol medium.
The growth mechanism of TiO2 nanotubes has been studied by many researchers.34,35 There are three processes, i.e., field-assisted oxidation at the metal–oxide interface, field-assisted dissolution at the oxide–electrolyte interface at the tube bottom and chemical dissolution at the tube mouth. The formation of nanotubes is a direct consequence of the competition between the electrochemical etching rate and the chemical dissolution rate. The former rate is determined by field-assisted oxidation and dissolution. The latter rate is determined by chemical dissolution. In our work, the high applied potential enhances the ion migration in the electrolyte and ion transport in the anodic barrier layer, resulting in a rapid electrochemical etching rate. On the other hand, the addition of lactic acid promotes the chemical dissolution rate. Therefore, the thin layer of oxides on top of the nanotube arrays formed in the first few tens of seconds is rapidly dissolved. At the same time, a balance of the electrochemical etching rate and the chemical dissolution rate can be attainable in a short time.
XRD analysis was used to confirm the crystal phases of the TiO2 nanotubes and the Pt-based catalysts. The results are shown in Fig. 2. After annealing at 400 °C, the as-prepared TiO2 nanotubes are in perfect anatase phase. The diffraction peaks are very sharp and correspond with the standard spectrum. For the Pt/C–TNTs-EG catalyst, the diffraction peaks of anatase TiO2 remain and are significant, which indicates that the synthesis process of the catalyst has no influence on the crystallization of the TiO2 nanotubes. In addition, the representative diffraction peaks of Pt are unclear due to the suppression effect of the TiO2 strong peaks. However, the diffraction peak at 2θ ≈ 40° is significantly broadened, which is assigned to the overlapping of the Pt (111) diffraction peak. Similarly, the diffraction peak of amorphous carbon at 2θ ≈ 26° is unclear and only a small heave indicates the existence of carbon. The unstable base of the Pt/C–TNTs-EG pattern is also evidence for the presence of carbon.
The morphology of the obtained samples was further determined by TEM as shown in Fig. 3a. The diameter of the nanotubes is about 110 nm and the length is inconsistent, which are consistent with the results of the SEM analysis. It is clearly seen that the distribution of Pt nanoparticles on the TNTs-EG is very uniform which is beneficial to the improvement of the catalyst. According to statistics based on the TEM results, the associated size distribution of the Pt nanoparticles was obtained and the mean size of the Pt nanoparticles was found to be 2.4 nm as shown in Fig. 3b. The high-resolution TEM (HRTEM) image obtained of the red box region in Fig. 3a shows that the lattice fringes can be coherently extended across the whole area, indicating that the sample has good crystallization. As shown in Fig. 3c, the d-spacings of 0.35 nm and 0.23 nm correspond respectively to the (101) plane of the anatase TiO2 and the (111) plane of the face-centered cubic Pt structure. Furthermore, the fast Fourier transform (FFT) pattern confirms that the TiO2 nanotubes have a regular single crystal structure as shown in Fig. 3d.
![]() | ||
Fig. 3 The TEM image (a), size distribution of the Pt nanoparticles (b), HRTEM image (c), and FFT pattern (d) of the Pt/C–TNTs-EG. |
The chemical state information of the Pt element in the Pt/C–TNTs-EG was analyzed by X-ray photoelectron spectroscopy (XPS), and the commercial Pt/C was used as a comparison. As shown in Fig. 4, the Pt 4f peak exhibits two associated peaks, representing the Pt 4f7/2 peak and the Pt 4f5/2 peak. Curve fitting of the Pt 4f peaks was carried out to gain the ratio of different valence and the deconvoluted results are shown in Table 1. The Pt(0) content of the Pt/C–TNTs-EG and the commercial Pt/C is similar, respectively 39.8% and 41.0%. However, the Pt(II) and Pt(IV) contents of the Pt/C–TNTs-EG and the commercial Pt/C are distinctly different. The Pt(II) content of the Pt/C–TNTs-EG is much higher than that of the commercial Pt/C while the Pt(IV) content of the Pt/C–TNTs-EG is much lower. During the electrochemical activation stage, Pt(II) is more easily converted to Pt(0) which could act as catalytic sites. Hence, the Pt/C–TNTs-EG would exhibit a higher catalytic performance than the commercial Pt/C. In addition, the binding energy of Pt(0) for the Pt/C–TNTs-EG shows a shift up of 0.18 eV in comparison with that of the commercial Pt/C, indicating that there is a metal–support interaction (MSI) between the Pt nanoparticles and titania. Specifically, there is a d–d-interelectronic bonding between the Pt and TiO2. The electron density on the Pt decreases due to the altervalent changes (Ti4+ ⇔ Ti3+) (eqn (1) and (2)).30 Therefore, the binding energy of Pt(0) shifts to a higher energy, and it can be deduced that the Ti3+ concentration should increase in the presence of Pt. In fact, it has been shown that the concentration of Ti3+ is higher in the presence of Pt than on the pure support.36
Ti(OH)4 + Pt → Ti(OH)3+ + Pt–OH + e− | (1) |
Ti(OH)3+ + 2H2O → Ti(OH)4 + H3O+ | (2) |
Sample | Pt species | Binding energy | Concentration |
---|---|---|---|
Pt/C–TNTs-EG | Pt(0) | 71.60 eV | 39.8% |
Pt(II) | 72.38 eV | 40.0% | |
Pt(IV) | 74.29 eV | 20.2% | |
Pt/C | Pt(0) | 71.42 eV | 41.0% |
Pt(II) | 72.42 eV | 27.7% | |
Pt(IV) | 74.12 eV | 31.3% |
![]() | ||
Fig. 5 Cyclic voltammograms of the catalysts in 0.5 mol L−1 H2SO4; the data for the commercial Pt/C were reproduced with permission.32 Copyright 2014, Elsevier. |
We then evaluated the electrocatalytic performance of the Pt/C–TNTs-EG for methanol electrooxidation (MOR). The MOR measurements were performed in an Ar-saturated solution of 0.5 mol L−1 H2SO4 containing 0.5 mol L−1 CH3OH at a scanning rate of 50 mV s−1 at 25 °C. Fig. 6 shows the MOR curves for the Pt/C–TNTs-EG and the commercial Pt/C. We can see that the MOR curves display two distinguishable oxidation peaks. The forward peak current density is an important indicator with regard to the methanol electrooxidation reaction. The forward peak current densities on the Pt/C–TNTs-EG and commercial Pt/C are 0.45 and 0.30 A mgPt−1, respectively, indicating that the activity of the Pt/C–TNTs-EG is very high, which is about 1.5 times higher than that of the commercial Pt/C. These results are consistent with the ESA results above.
![]() | ||
Fig. 6 Cyclic voltammograms of the catalysts in an acidic methanol medium; the data for the commercial Pt/C were reproduced with permission.32 Copyright 2014, Elsevier. |
The CO-stripping experiment was carried out as shown in Fig. 7. There is no significant difference in the onset potential of CO oxidation between the Pt/C–TNTs-EG and the commercial Pt/C. The reason may be as follows. The metal–support interaction induces an upshift of the Pt binding energy for the Pt/C–TNTs-EG, which simultaneously enhances the adsorption of OH and CO.30 The enhanced adsorption of OH can promote the CO oxidation while the enhanced adsorption of CO can block the CO oxidation. The opposite effect is possibly the reason that the onset potential of CO oxidation for the Pt/C–TNTs-EG does not shift. However, the metal–support interaction can weaken the adsorption of H, which is conducive to the efficient dehydrogenation on the Pt. In addition, the hydrogen spillover effect between Pt and TiO2 can also promote the dehydrogenation reaction on the Pt, creating more clean active sites on the Pt.37 Therefore, the electrochemically active specific surface area (ESA) and the activity of methanol oxidation for the Pt/C–TNTs-EG are higher than those of the commercial Pt/C.
We also evaluated the electrochemical durability of the Pt/C–TNTs-EG by using the constant potential test and cycling potential test in an Ar-saturated solution of 0.5 mol L−1 H2SO4 containing 0.5 mol L−1 CH3OH at a scanning rate of 50 mV s−1 at 25 °C. The amperometric i–t curves were obtained at a constant potential of 0.6 V vs. NHE for 3600 s as shown in Fig. 8. The final current densities after 3600 s were 22.2 and 7.6 mA mgPt−1, respectively, indicating that the activity of the Pt/C–TNTs-EG is higher than the commercial Pt/C. The durability of the catalysts can be evaluated by calculating the ratio of the final current to the maximum current. The retention rates of the mass current density are 19.3% and 11.0% for the Pt/C–TNTs-EG and commercial Pt/C catalysts, respectively, revealing that the durability of the Pt/C–TNTs-EG catalyst is higher than the commercial Pt/C.
The cycling durability behavior of the Pt/C–TNTs-EG and commercial Pt/C catalysts toward methanol electrooxidation was evaluated, and the results of the cycling aging tests are shown in Fig. 9. The cycling potential is between −0.64 V to 0.51 V vs. MSE. From the normalized peak current density as shown in Fig. 9c, the retention rates of the peak current density are 70.1% and 64.4% for the Pt/C–TNTs-EG and commercial Pt/C, respectively, after a 600 cycle test. The result indicates that the durability of the Pt/C–TNTs-EG is higher than that of the commercial Pt/C. In addition, it is noteworthy that the peak current density of the Pt/C–TNTs-EG is always much higher than that of the commercial Pt/C during the 600 cycle test as shown in Fig. 9d. The initial activity of the commercial Pt/C is equivalent to the activity of the Pt/C–TNTs-EG after the 600 cycle test. The result is exciting and hopeful for future applications.
![]() | ||
Fig. 9 Cycling aging test of the Pt/C–TNTs-EG (a) and the commercial Pt/C (b); the relationship of peak current density and cycle number (c and d); the data for the commercial Pt/C were reproduced with permission.32 Copyright 2014, Elsevier. |
This journal is © The Royal Society of Chemistry 2015 |