The effect of hydrothermal treatment time and level of carbon coating on the performance of PtRu/C catalysts in a direct methanol fuel cell

Abstract

A carbon-riveted PtRu/C catalyst of high stability has been prepared by in situ glucose carbonization using a hydrothermal method (GICH). Its mode of action and its practical application have been investigated by X-ray diffraction, high-resolution transmission electron microscopy, X-ray photoelectron spectroscopy, a single-fuel cell test, and by conventional electrochemical measurements. The single-fuel cell test has demonstrated that the GICH hydrothermal method has high applicational usefulness. After 100 h the maximum power density of a single cell using carbon-riveted PtRu/C as anode catalyst fell by only 12.0%, from 76.6 to 67.4 mW cm⁻², compared with 28.4%, from 73.2 to 52.4 mW cm⁻², for traditionally prepared PtRu/C. In addition, when the optimal hydrothermal treatment time was 4 h and the level of carbon coating was 9%, a carbon-riveted PtRu/C catalyst with a 3.5 nm carbon coating gave the best stability, with similar initial activity to traditionally prepared PtRu/C. The significantly increased stability of carbon-riveted PtRu/C may be attributed to two factors: (1) the anchoring effect of the carbon nanolayer formed during in situ glucose carbonization by the hydrothermal method; and (2) the increased content of Pt(0), Ru(0), sp³-hybridized carbon and the C–OR group composition, and the clear decrease in PtO₂ and RuO_xH_y following the carbon-riveting procedure.

---

1. Introduction

Compared with hydrogen proton exchange membrane fuel cells (PEMFCs), direct methanol fuel cells (DMFCs), with renewable liquid methanol as fuel, have a unique advantage since methanol is relatively safe during storage and transportation.^1–4 During methanol oxidation, pure platinum is poisoned by the adsorption of CO formed as an intermediate. The addition of ruthenium to the platinum improves the rate of methanol oxidation through a bifunctional mechanism.⁵ However, lack of stability of the PtRu/C catalyst has continued to hinder the commercialization of these systems for both stationary and transportation power application.

To improve the durability of catalysts for DMFCs or PEMFCs, extensive research has been carried out using a range of methods. Some researchers have focused on the preparation of Pt-based alloys,⁶ or their modification with other metals. For example, Pt–M (M = Cu, Co, Ni or Fe) nanowires have greater activity and stability than the corresponding pure Pt and Pt black nanowires.⁷ After 1000 cycles the potential between 0.2 and 1.0 V at a scan rate of 200 mV s⁻¹ in 1.0 mol L⁻¹ CH₃OH and 0.5 mol L⁻¹ H₂SO₄ solution, the current densities of the peaks for the commercial Pt black and Pt nanowires fall dramatically, with an initial activity of 0.20 and 0.47 A mg_Pt⁻¹, respectively. However, the current densities of Pt–M nanowires fall less on cycling, with an initial activity of 0.89, 0.82, 0.87 and 0.79 A mg_Pt⁻¹ for Pt–Cu, Pt–Ni, Pt–Co and Pt–Fe, respectively.

The second approach involves modification of the catalyst support. A variety of carbon-based materials,⁸ such as carbon nanotubes (CNT),^9–12 carbon nanofibers (CNF),¹³ or graphene,^14–16 as well as non-carbonaceous based materials, including titania,^17,18 indium oxides,¹⁹ silica,²⁰ tungsten oxide,^21–23 manganese dioxide²⁴ or graphitic carbon nitride (g-C₃N₄)^25–27 are attractive candidates for the catalyst support. Zhang and co-workers have synthesized and studied 3D-ordered mesoporous carbon sphere array-supported Pt nanoparticles (Pt/OMCS).²⁸ After 1000 cycles of the electrode potential between 0 and 1.3 V at a scan rate of 50 mV s⁻¹ in argon-purged 0.5 mol L⁻¹ H₂SO₄ solution at room temperature, the Pt/OMCS catalyst lost only 26% of its Pt ECSA, whereas that of the Pt/XC-72R and commercial Pt/C catalysts had decreased by 46% and 64%, respectively. The initial mass activity of Pt/OMCS, Pt/XC-72R and commercial Pt/C was 0.671, 0.118 and 0.145 A mg_Pt⁻¹, respectively. In addition there have also been reports focusing on migration and coalescence of metal particles by coating a porous shell such as SiO₂,²⁹ TiO₂,³⁰ or polybenzimidazole derivatives (PBIs)³¹ on the catalyst surface. Zhou and co-workers³² studied the stability of PtRu/CNT coated with a MnO₂ catalyst. After 2000 potential cycles in 1 mol L⁻¹ HClO₄ with 1 mol L⁻¹ CH₃OH, 55% and 30% activity remained of the initial activity of 0.48 and 0.44 A mg_Pt⁻¹ for MnO₂/PtRu/CNT and PtRu/CNT catalysts, respectively.

In our previous studies^33–36 we developed a method for increasing the stability of a PtRu/C catalyst by the in situ carbonization of glucose. Many highly stable catalysts were designed and prepared, such as Pt/C, Pt/TiO₂–C, Pt/MWCNTs–TiO₂, and Pt/MWCNTs–Al₂O₃. However, this method requires a high temperature around 400 °C, which leads to a remarkable increase in Pt nanoparticles. In addition, the carbon nanolayer formed by thermal methods may cover some of the active Pt sites. To solve this problem, carbon-riveted PtRu/C catalysts have been prepared by the hydrothermal method.³⁷ However, this work at present remains incomplete, and we do not yet have a thorough knowledge of the morphology and function of the carbon on the surface of the catalyst. The effect of the time of the hydrothermal treatment and the level of carbon coating on the performance of PtRu/C catalysts also requires further study.

In the present study we first confirmed that a carbonized carbon layer of about 2–5 nm was successfully coated on the surface of the PtRu/C catalyst. The nanolayers formed by the in situ carbonization of glucose greatly inhibited the migration and coalescence of PtRu nanoparticles on the support. However, the practical value of this method has been demonstrated by the single fuel cell life test. After a single cell life test extending over 100 h, the maximum power density of the carbon-riveted PtRu/C catalyst fell by only 12.0%, compared with 28.4% in the case of traditional PtRu/C.

2. Experimental

2.1. Preparation of the catalysts

All chemicals were analytical reagent quality. Hexachloroplatinic acid (H₂PtCl₆·6H₂O) and ruthenium chloride (RuCl₃) were purchased from the General Research Institute for Nonferrous Metals, Beijing, China. Vulcan XC-72 carbon black, of mean particle size around 20 nm, was purchased from Cabot, and 5 wt% Nafion® solutions were obtained from DuPont. The PtRu/C catalyst (metal loading 20 wt% supported by XC-72R, Cabot) was prepared by a microwave-assisted polyol process (MAPP),^22,23 and the atomic ratio of Pt/Ru was adjusted to an atomic ratio of 1 : 1, similar to that of the commercial catalyst.

Vulcan XC-72 carbon black (50 mg) was dispersed by ultrasonic treatment in 30 mL of a 4 : 1 v/v mixture of ethylene glycol (EG) and isopropyl alcohol in a 100 mL beaker for 1 h, to form a uniform carbon ink, and 0.0378 mol L⁻¹ H₂PtCl₆–EG (1.2 mL) and 0.02 mol L⁻¹ RuCl₃–EG (2.2 mL) were added and blended over 3 h. The pH was adjusted to 8 using a 1 mol L⁻¹ solution of NaOH in ethanol, and the suspension subjected to repeated 50 s microwave treatments in a microwave oven (Galanz Ltd, 800 W) under a flow of argon. After cooling to room temperature, the solution was adjusted to pH 2 using aqueous HNO₃ and stirred 12 h. Finally, the PtRu/C catalyst was filtered off, washed a number of times with ultrapure water (Millipore, 18.2 MΩ cm), dried for 3 h at 80 °C and stored under vacuum.

The PtRu/C prepared was riveted using a carbon nanolayer formed by the in situ carbonization of glucose using a hydrothermal method. 50 mg PtRu/C and the amount of glucose calculated to give a 6% mass ratio of carbon from the carbonization of glucose were dispersed in 250 mL ultrapure water in a beaker under ultrasonic treatment for 1 h. The suspension was then transferred to a reaction kettle (CJF-1L, Reflection Axe Industry Factory of China, Dalian) and argon gas fed in over 15 min to displace oxygen, for hydrothermal treatment at 160 °C for several hours. After cooling to room temperature, the carbon-riveted PtRu/C catalyst was washed, dried, and stored under vacuum.

The 40 wt% Pt/C and 40 wt% PtRu/C catalysts used in single fuel cells were prepared similarly. The amount of carbon riveted with 40 wt% PtRu/C was 9% and the hydrothermal treatment time 4 h.

2.2. Membrane electrode assembly (MEA) preparation

The 40 wt% Pt/C and the carbon-riveted 40 wt% PtRu/C were compared as anode catalysts in a single fuel cell. The anode catalyst comprised 40 wt% Pt/C and PtRu/C, and 5 wt% Nafion ionomer solution (DuPont, equivalent weight 1100) in isopropanol to form a homogeneous catalyst suspension. The cathodic catalyst was prepared similarly, with Pt/C, Nafion ionomer, and PTFE latex. The Nafion content in both anodic and cathodic catalyst layers was 20 wt%. The catalyst inks were painted onto the gas diffusion layers (GDLs) by brush, giving a metal loading of 2.5 mg cm⁻² on both electrodes. The anode and cathode GDL were each prepared by spray painting with 1 mg cm⁻² carbon black and 5% Nafion ionomer solution on carbon paper (Toray, TGPH 090). A DuPont Nafion 117 membrane was used as the solid electrolyte. Before applying to the electrodes, the Nafion membrane was pretreated by sequential immersion in boiling 3 wt% H₂O₂ solution in ultrapure water, followed by a boiling solution of 0.5 mol L⁻¹ H₂SO₄ in ultrapure water, each step occupying 1 h. The pretreated Nafion membranes sandwiched between the anode and cathode electrodes and the assemblies were hot pressed at 135 °C for 1.5 min under a loading of 100 kg cm⁻².

2.3. Physical characterization

2.3.1. X-ray diffraction (XRD). A D/max-RB powder diffractometer (Isuzu, Japan) was used to identify the XRD patterns of all the catalysts, using a Cu Kα X-ray source at 45 kV and 100 mA, scanning at 4° min⁻¹ with an angular resolution of 0.05°.

2.3.2. Transmission electron microscopy (TEM) and high-resolution transmission electron microscopy (HRTEM). TEM and HRTEM images of all the catalysts were characterized using a Tecnai G2 F30 field-emission TEM (FEI, Hillsboro, Oregon) at a spatial resolution of 0.17 nm. The samples were prepared by ultrasonically dispersing the catalyst powder in ethanol, and one drop of the suspension was deposited on a standard copper grid coated with carbon film and dried overnight. The applied voltage was 300 kV.

2.3.3. X-ray photoelectron spectroscopy (XPS). To determine the surface properties of the catalysts, XPS was carried out using a PHI model 5700 (Physical Electronics, Clanhassen, MN). Before XPS all samples were dried overnight in vacuo at 80 °C. The take-off angle of the sample to analyzer was 45° with the Al X-ray source at 250 W. Survey spectra were collected at pass energy (PE) 187.85 eV over a binding energy range between 0 and 1300 eV. High binding energy resolution multiplex data for the individual elements were collected at a PE of 29.55 eV. During all the XPS experiments the pressure inside the vacuum system was maintained at 1 × 10⁻⁹ Pa.

2.4. Electrochemical measurements

Electrochemical measurements were performed using a 650D potentiostat (CH Instruments, Texas) and a conventional three-electrode electrochemical cell. The counter electrode was a Pt sheet of 1 cm² plate and a Hg/Hg₂SO₄ electrode (−0.68 V relative to a reversible hydrogen electrode, RHE) was used as the reference electrode. The PtRu/C and carbon-riveted PtRu/C catalyst electrodes were used as working electrode. The catalyst ink was prepared by ultrasonically dispersing catalyst powder in the appropriate amount of ultrapure water. The catalyst ink of 5 μL was dropped onto a glassy carbon working electrode and dried for 15 min. 5 μL Nafion solution (5 wt%) was then spread over the surface of the electrode and dried in air. In all cases, the total loading of metal was 28 μg cm⁻².

Electrochemical measurements on the catalysts were carried out in a sealed glass cell containing 0.5 mol L⁻¹ H₂SO₄ and 0.5 mol L⁻¹ methanol solutions at 25 ± 1 °C. Highly purified argon gas was passed into the solution for 20 min to remove oxygen.

The stability of the catalyst was evaluated using an accelerated potential cycling test (APCT) conducted over the potential range 0.05–1.20 V (versus RHE) at a scanning rate of 50 mV s⁻¹. All potentials were reported relative to a reversible hydrogen electrode.

The electroactive specific surface area of the PtRu/C catalyst was determined by CO_ad-stripping voltammetry, assuming the formation of a monolayer of linearly adsorbed CO and the coulombic charge required for oxidation of CO_ad to be 420 mC cm⁻². The voltammetry was carried out at 25 °C in 0.5 mol L⁻¹ H₂SO₄ at a scanning rate of 50 mV s⁻¹.

Electrochemical tests on the MEAs were carried out by Fuel Cell Testing System (Scribner Associates Inc., Series 890E, Southern Pines, NC) using a single-cell technique (Electrochemistry Corp.). A methanol solution of 1.5 mol L⁻¹ was fed to the anode side at a flow rate of 3.0 mL min⁻¹. Pure oxygen was supplied to the cathode side at a rate of 200 mL min⁻¹ under ambient pressure. The cell was operated at 80 °C. The polarization curves and power density curves of the MEAs were plotted at intervals of operating time. Each point on the polarization curves and power density curves represented a steady-state performance achieved after about 3 min of continuous operation at a given voltage. Potential–time curves of the two single cells were plotted in galvanostatic mode at a current density of 150 mA cm⁻² for 100 h. To ensure that the electrolyte in the Nafion membrane and MEA electrode were moist enough to show high ionic conductivity, it was necessary to activate the MEA before the performance measurements. In our experiments, the single cells were conditioned with ultrapure water and oxygen at 80 °C for 5 h. Ultrapure water was then replaced with a methanol solution of 1.5 mol L⁻¹ for 20 h in galvanostatic mode at a current density of 30 mA cm⁻² prior to the acquisition of life data.

3. Results and discussion

Scheme 1 shows the fundamentals of in situ glucose carbonization by the hydrothermal method (GICH). Specifically, the mean size of the PtRu nanoparticles prepared became significant due to its coalescence by migration on the support surface during the APCT. However, that of the carbon-riveted PtRu nanoparticles grew slightly, since the existence of the carbon nanolayer on the surface of the support by in situ glucose carbonization inhibited the migration and coalescence of PtRu nanoparticles on the support.

Scheme 1 Fundamentals of in situ carbonization of glucose by the hydrothermal method (GICH).

For simplicity, the PtRu/C catalyst prepared traditionally is designated S-0, and carbon-riveted PtRu/C catalysts prepared at hydrothermal treatment times of 3, 4, and 5 h are designated S-3 h, S-4 h and S-5 h, respectively. Carbon-riveted PtRu/C catalysts prepared with coatings of 0%, 9%, and 12% are designated S-0%, S-9% and S-12%. In addition, the coating amount on S-3 h, S-4 h, and S-5 h catalysts was 6%. The hydrothermal treatment time for S-0%, S-9% and S-12% was the optimized time of 4 h.

3.1. Effect of hydrothermal treatment time on the performance of the PtRu/C catalyst

Firstly, we investigated the precise effect of hydrothermal treatment time on the performance of carbon-riveted PtRu/C catalyst. Fig. 1 shows the XRD patterns of S-0, S-3 h, S-4 h and S-5 h catalysts, respectively. The diffraction peaks at 26° are attributed to the hexagonal graphite structure (002) of the carbon black, and the 2θ values of the other four peaks correspond to the (111), (200), (220), and (311) crystal planes of crystalline face-centered cubic PtRu. It can clearly be seen that the crystallinity of PtRu nanoparticles increased with the hydrothermal treatment time.

Fig. 1 XRD patterns of (a) S-0, (b) S-3 h, (c) S-4 h and (d) S-5 h catalysts.

TEM images with the associated size distributions of S-0, S-3 h, S-4 h, and S-5 h catalysts before and after APCT are shown in Fig. 2, and more intuitive mean sizes of all samples are provided in Table 1. It can be clearly seen from Fig. 2 that the PtRu nanoparticles (S-0) were uniformly deposited on the surface of XC-72. With increasing hydrothermal time the PtRu nanoparticles showed some degree of aggregation. When the hydrothermal time was 5 h, the coalescence of PtRu particles became significant, as is clearly seen in Fig. 2(D-1 and D-2). The mean size of the PtRu nanoparticles of S-3 h, S-4 h, and S-5 h catalysts increased from the initial values of 2.1 to 2.3, 2.7 and 3.5 nm, respectively. It is reasonable that the coalescence of metal particles should increase with increasing hydrothermal time. With regard to the TEM images after APCT, the mean sizes of S-0, S-3 h, S-4 h, and S-5 h grew to 3.8, 3.1, 3.1 and 3.8 nm, increasing by 81%, 35%, 15%, and 10% in comparison to the size before APCT, respectively. Thus the process in Scheme 1 effectively anchored the crystallites and inhibited migration and agglomeration (coalescence) of the PtRu nanoparticles. The results of TEM are supported by the electrochemical measurements discussed below.

Fig. 2 TEM images and distribution of the particle sizes of PtRu/C catalysts after different hydrothermal treatment times: (A-1 and A-2) 0 h, (B-1 and B-2) 3 h, (C-1 and C-2) 4 h, (D-1 and D-2) 5 h before (A-1, B-1, C-1 and D-1), and after (A-2, B-2, C-2 and D-2) APTC.

Table 1 Mean sizes of the different PtRu/C catalysts

Sample	S-0 (A-1,2)	S-3 h (B-1,2)	S-4 h (C-1,2)	S-5 h (D-1,2)	S-0% (E-1,2)	S-9% (F-1,2)	S-12% (G-1,2)
Size before treatment (nm)	2.1	2.3	2.7	3.5	2.7	2.7	2.7
Size after treatment (nm)	3.8	3.1	3.1	3.8	4.0	3.0	3.1

---

The long-term stability of S-0, S-3 h, S-4 h, and S-5 h catalysts to methanol electro-oxidation was investigated by the continued CV cycles, as previously reported,³⁷ and the normalized peak current densities are presented in Fig. 3. It is particularly interesting that S-0 showed a sharp decline at 200 cycles and had lost almost 45% of its activity at 1000 cycles, compared to 40%, 22%, and 17% for S-3 h, S-4 h, and S-5 h, respectively. The sharp decline at 200 cycles may be due to the dissolution of Ru at high potential, leading to weakened CO tolerance.^38,39 Compared with conventional PtRu/C, the carbon-riveted PtRu/C catalysts (S-3 h, S-4 h, and S-5 h) showed ultra-high stability, and the methanol catalytic activity of S-3 h and S-4 h did not change significantly compared with S-0 before the APCT. In addition, the activity of the S-5 h catalyst became virtually constant after about 200 cycles. This behavior might be due to the fact that this had the largest metal particles of all the samples, about 3.5 nm before APCT, even larger than those of S-3 h and S-4 h after APTC. Undoubtedly, the larger size of catalyst particles gave greater stability. On the other hand, the initial activity of S-5 h catalyst was much lower than that of the other catalysts, therefore giving a relatively unobvious loss of activity subsequently.

Fig. 3 Cyclic voltammograms in 0.5 mol L⁻¹ H₂SO₄ and 0.5 mol L⁻¹ CH₃OH for PtRu/C at various hydrothermal treatment times: 0 h (A), 3 h (B), 4 h (C), 5 h (D), and their normalized peak current densities (E) during the APCT. Scanning rate: 50 mV s⁻¹; test temperature: 25 °C.

We consider that a shorter hydrothermal time (S-3 h) gave incomplete carbonization of the glucose. On the other hand, when the hydrothermal time was too long (S-5 h) the large size of the PtRu nanoparticles resulted in low catalytic activity. Combined with the results shown in Fig. 2, the CV results further demonstrate that carbon-riveted PtRu/C catalysts treated for 4 h gave the best performance in methanol electro-oxidation.

3.2. Effect of the level of carbon coating on the performance of PtRu/C catalyst

The effect of the level of coating on the performance of PtRu/C catalyst was further investigated. Fig. 4 shows the XRD patterns of carbon-riveted catalysts (S-0%, S-4 h, S-9%, and S-12%) with different coating amounts of 0%, 6%, 9%, and 12%, respectively, calculated before hydrothermal treatment. It is clearly seen that the crystallinity of the PtRu nanoparticles decreased with increasing levels of carbon coating. This was possibly due to the carbon nanolayer on the surface of the catalyst covering the metal particles to some extent, becoming thicker as the coating quantity increased.

Fig. 4 XRD patterns of (a) S-0%, (b) S-4 h, (c) S-9% and (d) S-12% catalysts.

TEM images with associated size distributions of the S-0%, S-9%, and S-12% catalysts before and after APCT are shown in Fig. 5. After similar 4 h hydrothermal treatment, these catalysts all had a similar nanoparticle size of 2.7 nm. With regard to the TEM images after APCT, it can be clearly seen that the mean sizes of three carbon-riveted PtRu/C catalysts (S-4 h, S-9% and S-12%) had grown to 3.1, 3.0 and 3.1 nm, respectively. However, the mean sizes of traditional PtRu/C (S-0%) nanoparticles grew from an initial size of 2.7 nm to 4.0 nm. The lower growth in size of the three carbon-riveted PtRu/C catalysts was due to the fact that the riveted carbon layer formed during the hydrothermal treatment effectively anchored the crystallites, inhibiting migration and agglomeration (coalescence) of the PtRu nanoparticles during APCT. In addition, although S-12% had the smallest size after APCT, it showed poor performance in methanol electro-oxidation. The reasons for this anomalous behavior are seen in Fig. 6.

Fig. 5 TEM images and distribution of particle sizes of PtRu/C catalysts with different amounts of carbon coating: (E-1 and E-2): 0%, (F-1 and F-2): 9%, (G-1 and G-2): 12%, before (E-1, F-1 and G-1), and after (E-2, F-2 and G-2) APTC.

Fig. 6 HRTEM images of (a) S-0%, (b) S-4 h, (c) S-9% and (d) S-12% catalysts.

Fig. 6 shows HRTEM images of S-0%, S-4 h, S-9%, and S-12% before APCT. The carbonized carbon layers of about 2.0, 3.5 and 5.1 nm for S-4 h, S-9%, and S-12%, respectively, can clearly be observed, although no obvious carbon layers can be seen for the S-0% sample in Fig. 6(a). It is not surprising that the carbon layers became thicker with increasing levels of coating. Combined with the results of CV from Fig. 7, S-9% catalyst, with a carbonized carbon layer of 3.5 nm, gave the best performance. For S-12% catalyst, the thicker carbonized carbon layers significantly covered the PtRu metal nanoparticles, leading to reduced catalytic activity. In addition, the similar catalytic activity of S-0%, S-4 h, and S-9% confirmed that the porous carbonized carbon layers on the catalyst surface were tending to cover the active catalytic sites.

Fig. 7 The CO_ad stripping voltammetry on traditional PtRu/C (S-0) and the carbon-riveted PtRu/C catalyst (S-9%). Scanning rate: 50 mV s⁻¹.

The fact that the porous carbonized carbon layer was tending to covering the active sites was confirmed by the CO_ad stripping voltammograms shown in Fig. 7. The electrochemical active specific surface areas (ESA) of S-0 and S-9% calculated by CO_ad stripping voltammetry were 89.1 and 86.3 m² g⁻¹ Pt, respectively. The slightly decreased ESA was due to the porous carbon layer over the active sites. The onset potential for oxidation of adsorbed CO on the S-9% catalyst was shifted to a lower electrode potential by 47.0 mV compared with traditional PtRu/C. The negative shift in potential originated from the facts that the size of PtRu nanoparticles increased⁴⁰ and the surface ratio of Pt and Ru became more suitable for electro-oxidation of the CO molecules adsorbed on the carbon-riveted PtRu/C catalyst after hydrothermal treatment.

Fig. 8 shows the long-term stability behavior of carbon-riveted catalysts with different amounts of coating, 0%, 6%, 9%, and 12%, and their normalized peak current densities. After 1000 APTC, S-0% catalyst showed a decline of 27.0% in current density, compared with 22.0%, 19.8%, and 17.2% for S-4 h, S-9%, and S-12%, respectively. With increasing carbon coating, the stability of these catalysts gradually increased. Clearly S-9%, with a carbon coating of 9%, showed the best catalytic activity and stability. This may be compared with the 44.3% current density decrement in S-0 shown in Fig. 3, which in S-9% with similar activity is only 19.8% after APCT of 1000 cycles, and 9% was therefore taken as the optimal carbon coating amount under our experimental conditions.

Fig. 8 Cyclic voltammograms in 0.5 mol L⁻¹ H₂SO₄ and 0.5 mol L⁻¹ CH₃OH for carbon-riveted PtRu/C of different coating amounts, 0%, 9%, 12%, and their normalized peak current densities during the APCT. Scanning rate: 50 mV s⁻¹; test temperature: 25 °C.

Deconvoluted Pt4f, Ru3p, C1s, and O1s peaks from the XPS analysis of S-0 and S-9% catalysts are shown in Fig. 9. The curves fitting the Pt4f, Ru3p, C1s, and O1s peaks of the X-ray photoelectron spectra for S-0 and S-9% catalysts were in agreement with our previous results.³³ The binding energies of all components, together with their relative intensities, are summarized in Tables 2–5. Not surprisingly, the content of Pt(0) increased by 12.53%, accompanied by a decrease of 9.11% in Pt(II), demonstrating that the carbon-riveted PtRu/C catalyst had greater stability relative to the traditional PtRu/C catalyst, due to the higher corrosion resistance of Pt(0).⁴¹ Consistent with the results of Pt4f, the content of Ru(0) increased from 21.99 to 39.66% following in situ glucose carbonization. The increased metallic Pt and Ru content of the S-9% sample can be attributed to the greater reducibility of glucose at high temperatures. The XPS results of C1s spectra show that the relative intensities of the oxygen-containing functional groups in the carbon-riveted PtRu/C catalyst were higher than those of traditional prepared PtRu/C. The oxygen-containing functional groups formed by the carbonization of glucose can effectively anchor and stabilize the metal NPs.⁴² In addition, it is also clearly seen in Table 3 that the sp³-C in the carbon-riveted PtRu/C catalyst was 8.49% greater than that in traditional PtRu/C, indicating that carbon resulting from the carbonization of glucose had greater stability than XC-72 carbon black. This may be another reason for the ultra-high stability of the carbon-riveted PtRu/C catalyst.

Fig. 9 Deconvoluted Pt4f (C and G), Ru3d (D and H), C1s (A and E) and O1s (B and F) peaks from XPS analysis of PtRu/C S-0 (A, B, C and D) catalyst traditionally prepared and riveted PtRu/C S-9% (E, F, G and H) catalyst.

Table 2 Results of the fits of the Pt4f spectra

Catalysts	Species	Orbital spin	Binding energy (eV)	Peak half width (eV)	Assignment	Relative content (%)
Traditional PtRu/C	Pt4f	4f7/2	71.71	1.69	Pt	30.95
		4f5/2	75.11	2.02	Pt	23.02
		4f7/2	72.88	1.64	PtO	10.85
		4f5/2	76.28	2.35	PtO	8.20
		4f7/2	74.46	3.50	PtO2	15.61
		4f5/2	77.86	3.08	PtO2	11.38
Riveted PtRu/C	Pt4f	4f7/2	71.79	1.71	Pt	38.11
		4f5/2	75.19	1.69	Pt	28.39
		4f7/2	73.25	1.38	PtO	5.88
		4f5/2	76.65	1.19	PtO	4.60
		4f7/2	74.52	2.37	PtO2	13.30
		4f5/2	77.92	2.81	PtO2	9.72

---

Table 3 Results of the fits of the Ru3p spectra

Catalysts	Species	Orbital spin	Binding energy (eV)	Peak half-width (eV)	Assignment	Relative content (%)
Traditional PtRu/C	Ru3p	3p1/2	461.68	3.60	Ru	21.79
		3p1/2	463.33	2.80	RuO2	43.59
		3p1/2	464.85	3.40	RuOxHy	34.62
Riveted PtRu/C	Ru3p	3p1/2	462.73	3.00	Ru	39.66
		3p1/2	464.19	2.64	RuO2	33.62
		3p1/2	466.42	3.25	RuOxHy	26.72

---

Table 4 Results of the fits of the C1s spectra

Catalysts	Species	Bond	Binding energy (eV)	Peak half width (eV)	Relative content (%)
Traditional PtRu/C	C1s	sp^2-C	284.54	1.36	57.46
		sp^3-C	285.20	1.18	19.98
		C–OR	286.10	2.02	18.07
		C=O	287.60	2.65	4.26
		COOR	288.74	1.73	0.22
		π	291.61	1.33	0
Riveted PtRu/C	C1s	sp^2-C	284.55	1.21	47.84
		sp^3-C	285.22	1.24	28.47
		C–OR	286.17	1.76	19.59
		C=O	287.60	1.66	1.59
		COOR	287.97	1.40	1.71
		π	288.94	1.42	0.80

---

Table 5 Results of the fits of the O1s spectra

Binding energy/eV	531.20 (–C=O)	532.26 (–OH)	533.39 (R–O–R)	534.42 (–COOH)	539.67 (H2O)
Traditional PtRu/C	43.61	27.07	17.29	9.02	3.10
Riveted PtRu/C	37.39	41.89	9.46	9.01	2.25

---

In order to confirm the stability of carbon-riveted catalyst in practical terms, the life tests of two single cells, which separately used 40 wt% traditional PtRu/C and 40 wt% carbon-riveted PtRu/C as anode catalysts, were carried out at a cell temperature of 80 °C at a high current density, 150 mA cm⁻². As shown in Fig. 10a, the majority of voltage losses occurred within the first few hours, and thereafter their decay became less significant. The initial rapid loss of performance may be attributed to the non-equilibrium state of ruthenium oxides. During the whole-life test, the cell voltages decreased with test time and there was a slow irrecoverable performance loss, which might be related to the degradation of catalysts, the solution of Nafion in the catalyst layers, and the ageing of the polymer–electrolyte membrane. Apart from the difference in anode catalyst in the two single fuel cells, the other elements were unchanged, and the reduced performance was therefore a direct reflection of the stability of the two catalysts. Compared with the voltage decay of 82.9 mV for the reference DMFC used in the traditional PtRu/C catalyst, in the carbon-riveted PtRu/C this was only about 30.0 mV after a constant current life test of 100 h. The carbon-riveted catalyst showed the higher stability.

Fig. 10 (a) Life tests of DMFC using two single cells with an apparent cross-sectional area of 5 cm² for different times. Anodic catalyst: 40 wt% traditional PtRu/C or 40 wt% carbon-riveted PtRu/C, metal loading 2.5 mg cm⁻². Cathodic catalyst: 40 wt% Pt/C, metal loading 2.5 mg cm⁻². (b) Performance of single DMFC using 40 wt% traditional PtRu/C, before and after different test times. (c) Performance of single DMFC using carbon-riveted 40 wt% PtRu/C before and after different test times. Operating conditions: 80 °C, 150 mA cm⁻². Anodic feed: 1.5 mol L⁻¹ CH₃OH solution at a flow rate of 3.0 mL min⁻¹. Cathodic feed: oxygen at ambient pressure and a flow rate of 200 mL min ⁻¹.

Cell performance before and after the life test was compared by polarization and power density curves for the two single cells. The cell performances show different degrees of decay with test time. The maximum power densities (MPD) before the life test were similar, about 73.2 and 76.6 mW cm⁻² for traditional and carbon-riveted PtRu/C catalysts, respectively. In the case of the cell using the traditional catalyst shown in Fig. 10(b), the MPD fell by 28.4% after a test time of 100 h. Under similar conditions, the DMFC using carbon-riveted PtRu/C as anode catalyst in Fig. 10(c) showed better stability, with a lower fall in MPD of 12.0%. In addition, the loss of power density at a cell voltage of 0.4 V was also lower than that of the reference DMFC, about 6.9 and 32.3% for traditional and carbon-riveted PtRu/C catalysts, respectively.

4. Conclusions

A carbonized layer of 2–5 nm has been successfully coated on the surface of a traditional PtRu/C catalyst. The optimal hydrothermal treatment time was 4 h, and shortage of hydrothermal time led to incomplete carbonization of the glucose. On the other hand, when the hydrothermal time was too long the large size of the PtRu nanoparticles gave inferior catalytic activity. When the carbon coating amount was 9%, the carbon-riveted PtRu/C catalyst with a 3.5 nm carbonized carbon layer gave the best performance. The accelerated potential cycling test and the life test of a single cell both showed that the carbon-riveted PtRu/C catalyst prepared for 4 h with carbon coating of 9% gave greater stability with similar initial activity compared to the traditional PtRu/C. After a 100 h life test, the maximum power density of a single fuel cell employing the carbon-riveted PtRu/C as anode catalyst fell by only 12.0%, from 76.6 to 67.4 mW cm⁻², compared with a 28.4% fall, from 73.2 to 52.4 mW cm⁻², for the traditional PtRu/C. The voltage decay of 82.9 mV for the DMFC using the traditional PtRu/C catalyst was almost twice as high as that (about 30.0 mV) using the carbon-riveted PtRu/C after a constant current life test of 100 h. The significantly increased stability obtained with carbon-riveted PtRu/C catalyst could be attributed to two factors: (1) the anchoring effect of the carbon nanolayer formed during in situ glucose carbonization by the hydrothermal method; and (2) the increasing content of Pt(0), Ru(0), sp³-hybridized carbon and C–OR groups, and the decreasing content of PtO₂ and RuO_xH_y after the carbon-riveting process.

Acknowledgements

This research was financially supported by the National Natural Science Foundation of China (Grant no. 21273058), the China Postdoctoral Science Foundation (Grant no. 2012M520731 and 2014T70350), Heilongjiang Postdoctoral Foundation (LBH-Z12089) and the Outstanding Subject Leaders of the Special Project Fund of Harbin in China (Grant no. 2012RFXXG99).

References

1. Z. Wang, G. Shi, J. Xia, F. Zhang, Y. Xia, Y. Li and L. Xia, Chin. J. Chem., 2013, 71, 1225–1238.
2. S. Sharma and B. G. Pollet, J. Power Sources, 2012, 208, 96–119.
3. M. K. Debe, Nature, 2012, 486, 43–51.
4. X. Zhao, M. Yin, L. Ma, L. Liang, C. Liu, J. Liao, T. Lu and W. Xing, Energy Environ. Sci., 2011, 4, 2736–2753.
5. L. La-Torre-Riveros, R. Guzman-Blas, A. E. Mendez-Torres, M. Prelas, D. A. Tryk and C. R. Cabrera, ACS Appl. Mater. Interfaces, 2012, 4, 1134–1147.
6. Y. Okawa, T. Masuda, H. Uehara, D. Matsumura, K. Tamura, Y. Nishihata and K. Uosaki, RSC Adv., 2013, 3, 15094.
7. X. F. Yu, D. S. Wang, Q. Peng and Y. D. Li, Chem.–Eur. J., 2013, 19, 233–239.
8. A. Santasalo-Aarnio, M. Borghei, I. V. Anoshkin, A. G. Nasibulin, E. I. Kauppinen, V. Ruiz and T. Kallio, Int. J. Hydrogen Energy, 2012, 37, 3415–3424.
9. X. L. Yang, X. Z. Liu, X. Y. Meng, X. Y. Wang, G. Li, C. Y. Shu, L. Jiang and C. R. Wang, J. Power Sources, 2013, 240, 536–543.
10. A. Y. Lo, N. Y. Yu, S. J. Huang, C. T. Hung, S. H. Liu, Z. B. Lei, C. T. Kuo and S. B. Liu, Diamond Relat. Mater., 2011, 20, 343–350.
11. D. P. He, C. Zeng, C. Xu, N. C. Cheng, H. G. Li, S. C. Mu and M. Pan, Langmuir, 2011, 27, 5582–5588.
12. L. Jiang, H. Fu, L. Wang, W. Zhou, B. Jiang and R. Wang, RSC Adv., 2014, 4, 51272–51279.
13. S. Kang, S. Lim, D. H. Peck, S. K. Kim, D. H. Jung, S. H. Hong, H. G. Jung and Y. Shul, Int. J. Hydrogen Energy, 2012, 37, 4685–4693.
14. K. Kakaei and M. Zhiani, J. Power Sources, 2013, 225, 356–363.
15. K. Ji, G. Chang, M. Oyama, X. Z. Shang, X. Liu and Y. B. He, Electrochim. Acta, 2012, 85, 84–89.
16. D. Bin, F. Ren, H. Wang, K. Zhang, B. Yang, C. Zhai, M. Zhu, P. Yang and Y. Du, RSC Adv., 2014, 4, 39612–39618.
17. W. Wang, H. Wang, J. Key, V. Linkov, S. Ji and R. F. Wang, Ionics, 2013, 19, 529–534.
18. K. W. Park, Y. W. Lee, J. K. Oh, D. Y. Kim, S. B. Han, A. R. Ko, S. J. Kim and H. S. Kim, J. Ind. Eng. Chem., 2011, 17, 696–699.
19. H. Chhina, S. Campbell and O. Kesler, J. Power Sources, 2006, 161, 893–900.
20. T. Zhu, C. Y. Du, C. T. Liu, G. P. Yin and P. F. Shi, Appl. Surf. Sci., 2011, 257, 2371–2376.
21. Y. Zhou, W. M. Liu, X. C. Hu, Y. Q. Chu and C. A. Ma, Acta Phys.-Chim. Sin., 2013, 29, 1487–1493.
22. Y. Zhou, Y. Q. Chu, W. M. Liu and C. A. Ma, Acta Phys.-Chim. Sin., 2013, 29, 287–292.
23. F. Wu, Y. H. Liu and C. Wu, Rare Met., 2010, 29, 255–260.
24. A. T. E. Vilian, M. Rajkumar, S.-M. Chen, C.-C. Hu, K. M. Boopathi and C.-W. Chu, RSC Adv., 2014, 4, 41387–41397.
25. H. Huang, S. Yang, R. Vajtai, X. Wang and P. M. Ajayan, Adv. Mater, 2014, 26, 5160–5165.
26. M. Kim, S. Hwang and J. S. Yu, J. Mater. Chem., 2007, 17, 1656–1659.
27. C.-Z. Li, Z.-B. Wang, X.-L. Sui, L.-M. Zhang, D.-M. Gu and S. Gu, J. Mater. Chem. A, 2014, 2, 20139–20146.
28. C. Zhang, L. Xu, N. Shan, T. Sun, J. Chen and Y. Yan, ACS Catal., 2014, 4, 1926–1930.
29. S. Takenaka, H. Matsumori, H. Matsune, E. Tanabe and M. Kishida, J. Electrochem. Soc., 2008, 155, B929–B936.
30. B. Y. Xia, S. Ding, H. B. Wu, X. Wang and X. Wen, RSC Adv., 2012, 2, 792–796.
31. T. Fujigaya and N. Nakashima, Adv. Mater., 2013, 25, 1666–1681.
32. C. Zhou, F. Peng, H. Wang, H. Yu, C. Peng and J. Yang, Electrochem. Commun., 2010, 12, 1210–1213.
33. Z. Z. Jiang, Z. B. Wang, D. M. Gu and E. S. Smotkin, Chem. Commun., 2010, 46, 6998–7000.
34. Z.-Z. Jiang, Z.-B. Wang, Y.-Y. Chu, D.-M. Gu and G.-P. Yin, Energy Environ. Sci., 2011, 4, 728–735.
35. Z.-Z. Jiang, Z.-B. Wang, Y.-Y. Chu, D.-M. Gu and G.-P. Yin, Energy Environ. Sci., 2011, 4, 2558–2566.
36. Z. Z. Jiang, Z. B. Wang, W. L. Qu, H. Rivera, D. M. Gu and G. P. Yin, Nanoscale, 2012, 4, 7411–7418.
37. Z.-B. Wang, C.-Z. Li, D.-M. Gu and G.-P. Yin, J. Power Sources, 2013, 238, 283–289.
38. K. Matsuoka, S. Sakamoto and A. Fukunaga, J. Power Sources, 2013, 238, 251–256.
39. Y. Sugawara, A. P. Yadav, A. Nishikata and T. Tsuru, J. Electrochem. Soc., 2008, 155, B897.
40. T. Saida, W. Sugimoto and Y. Takasu, Electrochim. Acta, 2010, 55, 857–864.
41. W. Li and A. M. Lane, Electrochem. Commun., 2009, 11, 1187–1190.
42. Y. J. Kuang, Y. Cui, Y. S. Zhang, Y. M. Yu, X. H. Zhang and J. H. Chen, Chem.–Eur. J., 2012, 18, 1522–1527.

---