Platinum nanoparticles with superacid-doped polyvinylpyrrolidone coated carbon nanotubes: electrocatalyst for oxygen reduction reaction in high-temperature proton exchange membrane fuel cell

Abstract

In order to improve the catalytic activity and durability of proton-exchange-membrane-fuel-cells (PEMFCs), Nafion-free Pt-based catalyst using the superacid-doped polymer coated multiwall carbon nanotubes (MWCNTs) was investigated. The modification and nano polymerization of MWCNTs were developed by polyvinylpyrrolidone (PVP). The following observations were made in the presence of polymer: better dispersion of MWCNTs, higher thermal stability of MWCNT/PVP than that of pristine MWCNT up to 450 °C as tested by thermal gravimetric analysis (TGA), homogeneous distribution of Pt without agglomeration as observed by transmission electron microscope (TEM), and not too much difference in Pt loading amount as analyzed by inductively coupled plasma mass spectrometry (ICP-MS). Modification of the electrocatalyst was determined by Raman spectroscopy, X-ray diffraction (XRD) patterns and TEM. Proton conductivity of the electrocatalysts was carried out by employing the super proton conductor phosphotungstic acid (PWA). Presence of PWA in the MWCNT/PVP/Pt as well as electrostatic interaction between PVP chains and PWA particles confirmed based on X-ray photoelectron spectroscopy (XPS). We developed a novel cross-linking between MWCNTs using PWA/PVP composite. The individual MWCNTs were cross-linked with each other to form a network structure after addition of PWA which modulates the electronic network structure, considerably enhanced the electrocatalytic activity toward the oxygen reduction reaction (ORR) in acidic media with the most efficient four-electron transfer process. The results showed that by doping of PWA into MWCNT/PVP/Pt, utilization efficiency of the catalyst was significantly improved. The findings have huge implications for electrocatalysts and high-temperature proton-exchange-membrane (HTPEM) fuel cell systems.

---

Introduction

A fuel cell is an electrochemical device that converts chemical energy directly into electricity through an electrochemical reaction. Among the various types of fuel cells the proton-exchange-membrane-fuel-cells (PEMFCs), with high power density and ability to quick start, have attracted great attention in recent decades. Fuel cell technology will include several renewable alternative energy sources and will compete with conventional energy sources (such as internal combustion engines or batteries).^1–3 Their applications are categorized into three power board areas: portable, transportation, and stationary power plants. The development of proton exchange membranes to use in the advanced electrochemical energy devices such as fuel cells, is a critical issue. Unfortunately, current industry standard perfluorosulfonic acid (PFSA) membrane is limited by high cost and methanol permeability. Although PFSA such as Nafion fulfills most of the requirements for PEM fuel cell applications, the large-scale application of PEMs based on Nafion is limited by some reasons such as environmental incompatibility, high cost, dependency of its proton conductivity on humidity and reduced conductivity at high temperature.⁴ On the other hand, the limitation of using low temperature prevents the access to higher power densities.⁵ The instability of carbon support materials, in particular for platinum (Pt)-based catalysts at temperatures above ∼80 °C and low thermal durability of Nafion are the most important challenges associated with commercialization of high-temperature polymer electrolyte membrane (HTPEM) fuel cells.⁶ The advantages of using high temperatures include the increasing of electrochemical reaction rate (which in turn reduces the need for nanocatalyst), facilitating water management due to using only one phase of water, reusing waste heat, and reducing carbon monoxide poisoning of nanocatalysts.^7,8 Therefore, the prospective production technology leans toward HTPEM.^9–11 Unfortunately, carbon materials (Vulcan XC-72R and Ketjen) are not enough corrosion-resistant, so they are unlikely to meet automotive applications.¹² With respect to the mentioned issues and challenges, it becomes necessary to use alternative support materials. They should have some basic properties as follows: high corrosion-resistance in working conditions (under both dry and humid air), high surface area, good electrical and thermal conductivity, enough amount to disperse the precursor, as well as high electrochemical stability.⁶ In addition, strong interaction between supporting materials with catalytic metals has greatly improved their catalytic activity and durability.^10,13 Considerable efforts have been made to explore non-carbon support materials in the past decades.¹⁴ The developed supporting materials include metals,^15–18 nitrides,¹⁹ carbides,²⁰ mesoporous silicas,^21,22 conducting polymers^23–25 as well as metal oxides.^26–28 Composition of conductive polymers with carbon nanotubes (polymer-carbon nanotube supporter) can be used as catalyst-support materials,²⁹ which benefit from ionic conductivity, porous structure, high stability and better processability under ambient conditions.²⁹ In addition, to having a better distribution of catalyst nanoparticles, a good polymer supporter could also increase the reaction surfaces and facilitate the process of oxidation and reduction. Effective dispersion of carbon nanotubes has been reached by polymer wrapping around carbon nanotube and non-covalent interactions between them.³⁰ A great effort has been put in aromatic polymers such as polybenzimidazole (PBI) so that they can wrap around carbon nanotubes.^31–37 It is commonly recognized that the proton conductivity plays a critical role in membrane and electrocatalyst performance. Improving proton transport properties is common approach to achieve enhanced performance for large-scale applications. Furthermore, the slow kinetics of the oxygen reduction reaction (ORR) still remains one of the key reactions in fuel cells due to the sluggish electrochemical kinetics and large overpotential (or polarization), that strongly reduce the efficiency of the PEMFC.³⁸ Thus, investigation of the ORR remains the main contributor of PEMFC research.^39–41 Electron reaction can proceed accordingly to two pathways, one is the four-electron reduction of O₂ to H₂O pathway is likely to be the dominant mechanism, and the other is the production of hydrogen peroxide through a two-electron pathway. The ORR process is highly preferred by a four-electron transfer pathway.^42,43 Heteropolyacids (HPAs) are one of the most attractive inorganic modifiers to be highly conductive and thermally stable in their crystalline forms.^44,45 Heteropoly catalysts are used in solution as well as in the solid state.^46,47 HPAs have grown interestingly, due to their potential as composite membranes for higher temperature fuel cell⁴⁸ and their applications as the proton conducting component in a fuel cell membrane.⁴⁹ Also, Pt nanoparticles deposited on glassy carbon were modified with ultra-thin films of Keggin type heteropoly acids of molybdenum or tungsten for the reduction of oxygen in polymer electrolyte.⁵⁰ HPAs have the highest proton conductivity and can be used as strong solid-acid catalysts and proton-conducting materials.⁴⁵ Conductivity of HPAs is strictly linked to the number of water molecules per Keggin unit (KU). The Keggin [XM¹²O^4O]ⁿ⁻ structure consists of one XO₄ tetrahedral (X = P, S, Mo), surrounded by 12 octahedral MO₆ (M = Mo, W, V) connected with each other by the neighbouring oxygen atoms.⁵¹ The HPAs offer an excellent opportunity for creating a nanomaterial that can be incorporated in nanocomposite layers as a catalyst, but HPAs are water-soluble and must be immobilized to form practical materials with high proton conductivity to use in proton exchange membrane fuel cells. The strongest acid among the HPA is phosphotungstic acid (PWA) in which the central atom (P) is in a tetrahedral environment surrounded by 12 octahedral of WO₆. The oxygen atoms are all shared between tungsten atoms, except for 12 terminal oxygen atoms attached to only one W atom.^52,53 In addition, some PEM membranes were produced with low cost, non-conductive polymers by doping the phosphotungstic acid (PWA).^54,55 In addition, some polymers such as polyvinylpyrrolidone (PVP)³⁰ and poly(acrylic acid) (PAA),⁵⁶ are non-fluorinated polymers, which have a great potential for the construction of the electrocatalyst. They can also be used instead of expensive polymers such as PBI in fuel cell application. However, the leakage of PWA due to its great water-solubility is a drawback to use and commercialization and would decay the conductivity of the PEM gradually.^57,58 Thus, it is a critical challenge to immobilize PWA particles in a PEM to reduce its leakage and improve the stability. Considering the high electron transport, unique atomic structure and flexibility, MWCNTs are employed as ideal carbon support nanostructures for polymeric nanocomposites. However, these PWA will be desorbed from MWCNTs in the process of rinse during experiments, because they were deposited onto MWCNTs by weak molecular interaction.⁵⁹ In this study, we newly designed and developed a novel PVP nanocomposite in which MWCNTs, PWA and Pt nanoparticles are employed as a carbon supporter, proton transporter and redox catalyst, respectively. PVP can be used to immobilize negative PWA onto MWCNTs by electrostatic interaction. In contrast to net MWCNTs, MWCNT/PVP hybrids can be easily dispersed in water. There is evident that the PVP functioned not only as the wrapping polymer for the MWCNTs, but also as a Pt and PWA absorber through the coordination with the nitrogen on the PVP. PVP can be used as a linker to prepare MWCNT/PWA/Pt hybrids based on electrostatic interaction and N atoms of PVP can accept proton from PWA to charge positive, and thus loads Pt and negative PWA into MWCNTs. We have chosen PVP for four reasons (a) a Pt adsorbing material, (b) a multiwall carbon nanotube (MWCNT)-solubilizing material, (c) immobilizing of PWA due to electrostatic interaction and (d) a proton conductor. The ORR onset potential, mass activity, current density, the number of electrons, and kinetic current density was improved due to the strong PWA effect. We expect that this developed method can be extended in preparation of other hybrid nanostructures with high electrochemical performance.

Experimental

Materials

Multi-walled carbon nanotubes (MWCNTs, outer diameter 5–15 nm, average length ∼50 μm, specific surface area ∼233 m² g⁻¹ and purity > 95 wt%) were purchased from Nanosany. N,N-Dimethylacetamide (DMAc), polyvinylpyrrolidone (PVP, molecular weight 55 000 g mol⁻¹) and phosphotungstic acid (PWA) were supplied by Sigma. The other chemicals were purchased from Merck.

Synthetic procedures

Preparation of MWCNT/PVP nanocomposites. MWCNT/PVP was prepared based on the previous literature.⁶⁰ Briefly, PVP polymer and MWCNT were dispersed in deionized (DI) water (PVP/MWCNTs = 20 : 100 (w/w), MWCNTs/water = 100 mg/100 ml). They were stirred for 12 h at 60 °C, and then deagglomerated by sonication with a Wise Clean (2000 W) at an amplitude of 50% for 30 min in water bath at room temperature in order to obtain a black suspension of MWCNTs. This solution was ultracentrifuged (USA Beckman, IBB Institute) at 39 000 rpm (145 000 × g) for 1 h to remove unbound PVP. The sedimented MWCNT/PVP collected and subsequently was dispersed again in DI water by Misonix sonicator 3000 at a power of 6 W for 30 min. Centrifugation separation and collection of dispersed carbon nanotubes were performed 6 times.

Pt loading on the MWCNT and MWCNT/PVP composite. Aqueous solution of MWCNT/PVP and platinum salts H₂PtCl₆ (50 wt%) were refluxed in ethylene glycol solution under argon atmosphere at 120 °C for 24 h until Pt nanoparticles were completely reduced. The composition of MWCNT/PVP/Pt was centrifuged 39 000 rpm (145 000 × g) for 30 min and washed several times. Synthesis of MWCNT/Pt involved all above without PVP. The calcination of MWCNT/PVP/Pt and MWCNT/Pt was processed under argon atmosphere flow at 150 °C for 1 h.^60,61

PWA loading on MWCNT/PVP/Pt composite. MWCNT/PVP/Pt was uniformly dispersed in dissolved in DMAc (2 mg ml⁻¹), then an appropriate solution of PWA in DMAc (0.1 mg ml⁻¹) was added to the catalyst (MWCNT/PVP/Pt). The mixture (MWCNT/PVP/PWA/Pt) was sonicated at 50% of sonication power (amplitude) intensity for 1 minute. The results indicated that amount of PWA was largely dependent on the polymer content of catalyst. The amount of the added PWA was about 0.05 wt% of the total catalyst weight in this work.

Characterizations

Transmission electron microscope (TEM, Zeiss-EM10C-80 KV) was used to observe MWCNTs. Aqueous suspensions of MWCNTs were prepared and deposited on holey carbon coated grid Cu Mesh 300. Dispersion of MWCNT was performed by ultrasonic disperser (Misonix-S3000). Raman analysis (BRUKER, Germany, SENTERRA, Spectral Resolution: <3 cm⁻¹) and thermal gravimetric analysis in nitrogen atmosphere from 25 to 800 °C (NETZSCH STA 409) were used to characterize and evaluation of the carbon nanotubes and polymer stability. X-ray diffraction patterns (XRD, Co-Kα radiation source, Philips, X'Pert MPD) were studied in order to investigate the structures. X-ray photoelectron spectroscopy (XPS) were studied by “Bestec, Germany” located in Sharif University of Technology. Electrochemical characterization and CVs were measured by VSP-300 Multichannel potentiostat/galvanostat/EIS (Bio-Logic Science Instruments). Scan rate was 50 mV s⁻¹ from −0.25 to 1.2 V vs. Ag/AgCl and 20 °C. Electrolyte solution (1.0 M H₂SO₄) was deaerated with nitrogen. A three-electrode configuration was used consisting of a polished glassy carbon electrode (geometric area of 0.1963 cm²) as the working electrode, and Pt wire as the counter electrode. 20 μl of MWCNT/PVP/Pt dissolved in DMAc (2 mg ml⁻¹) was placed on a glassy carbon electrode. Finally, PWA was examined for high-performance proton exchange electrocatalyst. PWA (about 0.05% wt of catalyst) was mixed with 20 μl of MWCNT/PVP/Pt and blended before placing on glassy carbon.

Results and discussion

Raman spectroscopy

The analysis results of Raman spectra of PVP, H₂PtCl₆ and H₂PtCl₆/PVP obtained by mixing the aqueous solutions of H₂PtCl₆ and PVP, are shown in Fig. 1(a and b). The 400–700 cm⁻¹ region is suggested to be attributed to the Pt bands. Several peaks are observed in Raman spectra at 655, 632, 594, 495 and 444 cm⁻¹ for H₂PtCl₆/PVP. Band assignments are as follows: the peaks at 655, 632 and 594 cm⁻¹ originate from the reaction between amide group of the pyrrolidone rings of PVP and Pt nanoparticles (NC=O–Pt).^62,63 A red shift frequency is seen in Fig. 1(b) about 37 cm⁻¹ which is considered to be a strong reaction between the pyrrolidone rings and Pt nanoparticles. The peaks at 495 and 444 cm⁻¹ could be attributed to Pt–O and Pt–N stretching mode of pyrrolidone rings of PVP.

Fig. 1 Raman spectra of H₂PtCl₆, PVP and H₂PtCl₆/PVP mixtures (a and b) and MWCNT, MWCNT/PVP, MWCNT/PVP/Pt, MWCNT/Pt (c).

To gain more insight, Raman spectra of MWCNT/Pt with and without PVP were also investigated. As can be seen in Fig. 1(c), when Pt nanoparticles were reduced on the surface of the MWCNTs, the scattering intensity near radial breathing mode (RBM) was significantly augmented. Meanwhile, broad bands and Raman intensity were increased from approximately 300 to 400 cm⁻¹. PVP was found to be effective on the decrease of the vibration of Pt nanoparticles on the surface of MWCNTs. In this way, some nanoparticles were embedded in PVP polymer matrix in MWCNT/PVP/Pt. Therefore, Raman intensity decreased at 557 cm⁻¹ compared with MWCNT/Pt. Consequently, Raman intensity in MWCNT/Pt is higher than that of MWCNT/PVP/Pt within 557 to 600 cm⁻¹. The main parameters for precipitation of mixed materials have been shown by Raman spectroscopy. Consequently, Pt nanoparticle-adsorbed MWCNTs by polymer were uniformly dispersed on the MWCNTs surface and form new reinforced MWCNT/Pt nanocomposites.

Thermal gravimetric analysis (TGA)

The thermal gravimetric analysis (TGA) of the MWCNT/PVP composite after centrifugation followed by vigorous washing with distillated water to remove unbound PVP was performed and related curve obtained, after 450 °C corresponds to the thermal degradation of the wrapped PVP and just MWCNTs was remained (PVP was completely decomposed at about 450 °C). The MWCNT/PVP composite has a ∼11% weight loss due to PVP decomposition within 100 to 450 °C. Thus we can conclude that the first 11% weight loss that appeared at temperatures above 450 °C is attributable to the thermal oxidation of PVP and the remaining 91%-weight loss that appeared after 450 °C is attributed to the MWCNTs (Fig. 2). MWCNT/PVP composition is highly soluble in aqueous solution and a stable complex. As shown in Fig. 2, just 10 w% of electrocatalyst (MWCNT/PVP/Pt) was decomposed up to 400 °C. PVP modified MWCNT is more thermally stable than pristine MWCNT up to 450 °C.

Fig. 2 TGA analysis of: MWCNT, MWCNT/PVP, MWCNT/PVP/Pt, MWCNT/Pt and PVP.

Transmission electron microscopy (TEM)

As TEM images in Fig. 3(a and b) show, the average diameter size of MWCNTs is about 8–20 nm. Interestingly, PVP is wrapped around the surface of MWCNTs (Fig. 3c–f). Presence of PVP on MWCNTs surface makes them to repel each other in water by steric stabilization. It is observed that over the time, some metal nanoparticles are achieved by branched polymer arms of the PVP around the nanotubes (see Fig. 3(g)). In addition to being a better distribution of catalyst nanoparticles, the reaction surfaces can be progressively increased and facilitate the process of oxidation and reduction. The results suggest that in the presence of PVP, distribution of nanoparticles increases significantly (see Fig. 3g and h). The average particle size of Pt nanoparticles on MWNT/PVP/Pt and MWNT/Pt are determined to 4.2 nm and 4.5 nm, respectively by TEM observation shown in Fig. S1 (see ESI†). In a polymer-free synthesis, the Pt nanoparticles aggregated. The prepared Pt nanoparticles were not uniformly dispersed on the MWCNTs surface and there are many MWCNTs without Pt loading (see Fig. 3i and j).

Fig. 3 TEM images of MWCNTs (a and b), MWCNT/PVP (c–f), MWCNT/PVP/Pt (g and h) and MWCNT/Pt (i and j).

Formation of Pt nanoparticles on MWCNTs were also characterized by X-ray diffraction, and several peaks due to Pt face centered cubic structure are seen (see ESI, Fig. S2†). The average diameter of Pt nanoparticles on MWCNT/Pt, MWCNT/PVP/Pt, and MWCNT/Pt/PVP/PWA were measured to be 4.6, 4.6 and 5.1 nm respectively, by using the Debye Scherrer formula. Pt loading amount was determined by ICP about 50 w% for MWCNT/PVP/Pt and 53 w% for MWCNT/Pt. Transmission electron microscopy images of PWA doped MWCNT/PVP/Pt are shown in Fig. 4, primary nano-sized black of PWA particles can aggregate to form larger secondary clusters with average size 35 nm is clearly distinguished from MWCNT walls. It has been shown that the adsorption of PVP polymer on the outer wall of MWCNT is due to their π–π interactions. Moreover, due to high electrostatic attractive forces between PVP and PWA a physical cross-linking into the MWCNT network structures forms. The novel cross-linker was prepared by blend the PWA/PVP (1 : 200 w/w). Also cross-links can significantly improve the mechanical properties of structure. Interactions between oxygen atoms of PWA, which cover the periphery of PWA, and PVP chains, facilitate proton conductivity.

Fig. 4 TEM images of MWCNT/PVP/PWA/Pt (a and b).

X-ray photoelectron spectroscopy (XPS) analysis

X-ray photoelectron spectroscopy (XPS) was employed to investigate surface chemical elements and determine presence of PWA in MWCNT/PVP/PWA/Pt hybrids. The XPS profile of MWCNT/PVP/Pt and MWCNT/PVP/PWA/Pt hybrids was shown in Fig. 5. As shown in Fig. 5A, C 1s spectrum has its main features at 284.6 eV which originated from sp² carbon atoms of MWCNTs. W 4f, W 4d_3/2 and W 4d_5/2 signals arises in the 36.3, 248.6 and 262 eV, respectively originate from PWA indicating present of PWA in MWCNT/PVP/PWA/Pt hybrids (Fig. 5A). Pt 4f region exhibited doublet peaks at 71.1 and 74.3 eV, which are attributable to Pt⁰ 4f_7/2 and Pt⁰ 4f_5/2, respectively, indicating that the dominant valence of the Pt is zero as well as different peaks at 315 and 332 eV show presence of Pt in MWCNT/PVP/PWA/Pt and MWCNT/PVP/Pt hybrids. The O 1s signal of MWCNTs is dominant at 533 eV. Binding energy of O in MWCNT/PVP/Pt and MWCNT/PVP/PWA/Pt are almost identical, but after addition of PWA onto MWCNT/PVP/Pt hybrids, the O 1s signal of MWCNT/PVP/PWA/Pt hybrids intensified. This intensification is mainly derived from the group W–O of PWA, which accounted for the intensification of O 1s signal at XPS curve. Thus, presence of W peaks and intensification of oxygen peak shows presence of PWA in MWCNT/PVP/PWA/Pt hybrids. The wrapping of the PVP on the MWCNTs even after the Pt loading was confirmed by the presence of the N peaks at 395.3 and 398.46 eV corresponding to the double and single-bonded nitrogen, respectively. The binding energy of N peaks has a positive shift to 396.3 and 400.36 eV after addition of PWA, this shift guaranties the interaction between PVP and PWA, thus, it can be confirmed that the N atom in PVP interacts with the proton, which results PWA/PVP acts as a cross linker between MWCNTs and the formation of monolithic network (Fig. 4b).

Fig. 5 XPS survey spectra. (A) Recorded for (a) MWCNT/PVP/PWA/Pt, (b) MWCNT/PVP/Pt. (B) W 4f, (C) Pt 4f, (D) O 1s and (E) N 1s spectrum.

Cyclic voltammetry (CV)

Cyclic Voltammetry (CV) method was carried out to survey the electrochemical behaviour of catalysts. As shown in Fig. 6, the peaks associated with electrochemical adsorption/desorption of hydrogen, oxygen and water on platinum surfaces (Pt oxidation and reduction) can be seen at the potential range selected from −0.25 volt to 1.2 volt vs. KCl–calomel electrode in 1 M H₂SO₄. Three reduction peaks from −0.174 to 0.156 V correspond to H₂ electroadsorption on Pt(100), Pt(111) and Pt(110) crystal surfaces and three oxidation peaks correspond to H₂ electrodesorption on Pt(100), Pt(111) and Pt(110) crystal surfaces, respectively. The peak from 0.556 to 0.956 V is attributed to Pt oxidation in the positive-going scan, and reduction of Pt oxide in the negative-going scan. Based on our hypothesis, phosphotungstic acid (H₃PW₁₂O₄₀, PWA)-doped polymer wrapping around the MWCNT could be employed to enhance the proton conductivity of the obtained structures. As a consequence of PWA doping, Pt nanoparticles trapped in polymer could participate in reactions.

Fig. 6 CVs of nanocatalyst; 10th cycle (a), 100th cycle (b), 500th (c), electrochemical surface area for nanocatalysts (d); 20 μL of nanocatalyst dissolved in DMAc (2 mg ml⁻¹) was placed on a glassy carbon electrode (0.107 mg cm⁻² Pt for MWCNT/Pt and 0.101 mg cm⁻² Pt for MWCNT/PVP/Pt). 1 μL of PWA in DMAc (0.1 mg ml⁻¹) was mixed with 20 μL of nanocatalysts (MWCNT/Pt and MWCNT/PVP/Pt) for MWCNT/PVP/PWA/Pt. 20 μL of nanocatalyst dissolved in DMAc (2 mg ml⁻¹) with proper solution of Nafion ionomer (15 w%) was mixed for MWCNT/Pt/Nafion. CV was measured at the scan rate of 50 mV s⁻¹ at 20 °C in 1.0 M H₂SO₄ solution and deoxygenation with N₂ gas.

CV can also be used to determine the electrochemical surface area (ECSA, m² g⁻¹ catalyst). The ECSA can be estimated according to equation [eqn (1)]:⁶⁴

where Q_H is calculated as the average value coulombic charge (mC) exchange during the electro-absorption and electro-desorption of H₂ on Pt surface (from −0.174 to 0.156 V). L_Pt is amount of platinum loaded on the electrode (g) and Q_f is 0.21 mC cm⁻². As can be seen in Fig. 6(d), ECSA for MWCNT/PVP/PWA/Pt was calculated to be about 70 m² g⁻¹. It is higher than the previously reported value 51.6 m² g⁻¹ of Pt for MWCNT/Py/PBI/Pt, and 55.8 m² g⁻¹ of Pt for CB/Pt.³⁵ The highly efficient and homogeneous loading of Pt nanoparticles onto the MWCNTs was realized by the PVP wrapping on the MWCNTs as shown in Fig. 3 and 4. The results confirmed that the PVP can be served as an effective polymer linker to load the Pt and negative PWA onto the surfaces of MWCNTs by electrostatic adsorption. During the reaction, the protons of PWA molecule jump to N atoms of PVP, therefore, the proton pathway is also formed by adjacent PWA particles and the side chain N-heterocycle moieties of PVP. Therefore, there should be two kinds of proton transferring pathways in the MWCNT/PVP/PWA/Pt hybrid catalyst: (i) the proton can easily transfer from a PWA to another adjacent one directly when they are close,⁶⁵ (ii) the proton can transfer to PWA through the N-heterocycle assisted proton movement.⁶⁶ As a result, fast proton transfer occurs in the MWCNT/PVP/PWA/Pt hybrid catalyst. The high conductivity of the MWCNT/PVP/PWA/Pt/hybrid catalyst and lower activation energy, suggest a new nanostructured catalyst with the great potential application as a novel PEM. The electrostatic interaction and mechanism of electron and proton transport and their pathways are schematically shown in Fig. 7.

Fig. 7 (a) Electrostatic interaction between PWA molecule and PVP chains. (b) Mechanism and pathways of electron and proton transport and inset is TEM image of MWCNT/PVP/PWA/Pt.

Pt utilization efficiency of catalysts

The oxygen reduction reaction (ORR) is also the most important reaction in life processes. Pt is the most efficient and extensively discussed as electrocatalyst for ORR. Utilization efficiency (η_Pt) is a key parameter to describe the catalyst performance and is calculated by dividing ESCA by the chemical surface area (CSA), (ECSA/CSA). The method for measuring the surface area (CSA) is described by the following equation:where ρ is the density of Pt (21.09 g cm⁻³) and d is the mean diameter of the Pt nanoparticles obtained from TEM. The Pt utilization efficiency η of the MWNT/PVP/Pt was determined to be as high as 90%, while that of the MWNT/PVP/PWA/Pt was 98%. The utilization of the commercial catalyst (CB/Pt) is reported to be 54.8% (ref. 67) as shown in Table 1. The observed higher utilization efficiency of the MWNT/PVP/PWA/Pt is explained by the formation of the network structure of the catalyst formed by the PWA/PVP adsorption onto MWNTs.

Table 1 Comparison of Pt size, ECSA, CSA, and Pt utilization of various catalysts

Catalysts	Pt loading (%) from ICP	Pt size from (TEM) (nm)	ECSA from CV (m^2 g^−1)	CSA (m^2 g^−1)	ηPt (%)
MWCNT/PVP/Pt	50	4.2	60.91	67	90
MWCNT/PVP/PWA/Pt	50	4.2	69.9	67	98
MWCNT/Pt	53	4.5	52.78	63	83
MWCNT/Pt/PWA	53	4.5	47.26	63	75
Pt/C (E-TEK)	20	2.8	55.7	101.6	54.8

---

Electrocatalytic oxygen reduction reaction

To further evaluate the effect of the phosphotungstic acid on catalytic performance in the oxygen reduction reaction, the reaction kinetics carried out by using a glassy carbon rotating-disk electrode in O₂-saturated 0.5 M H₂SO₄ solution at different rotation speeds (from 100 to 4000 rpm) in which the current density increases with increasing rotation speeds. Polarization curves for the ORR on MWCNT/PVP/Pt and MWCNT/PVP/PWA/Pt catalysts are displayed in Fig. 8(a and b).

Fig. 8 Polarization curves for the ORR on MWCNT/PVP/Pt and MWCNT/PVP/PWA/Pt catalysts was measured by rotating-disk electrode in O₂-saturated 0.5 M H₂SO₄ solution at different rotation speeds (from 100 to 4000 rpm). (a and b) Corresponding Koutecky–Levich plots (c and d) and compares the rotating-disk voltammograms before and after PWA for ORR at the same rotating speed of 1000 rpm (e), mass and area-specific activity for MWCNT/PVP/Pt and MWCNT/PVP/PWA/Pt at 0.45 V (f).

The corresponding Koutecky–Levich plots (J⁻¹ vs. ω^−1/2) (Fig. 8(c and d)) exhibit good linearity, and the slopes remain approximately constant over the potential range of 0.2 to 0.55, thus suggesting the first-order reaction kinetics with respect to dissolved O₂ and the consistent electron transfer for oxygen reduction at different electrode potentials.⁶⁸ The kinetic parameters can be analyzed with the Koutecky–Levich equations [eqn (3)–(5)]:

(J⁻¹) = (J_L)⁻¹ + (J_k)⁻¹ = (Bw^1/2)⁻¹ + (J_k)⁻¹ (3)

B = nFC_OD_O^2/3ν^−1/6 (4)

J_k = nFC_O (5)

where J is the measured current density, J_K and J_L are the kinetic and diffusion limiting current densities, respectively, ω is the electrode rotation rate, n is the overall number of electrons transferred in oxygen reduction, F is the Faraday constant, C_O is the bulk concentration of O₂ dissolved in the electrolyte, D_O is the diffusion coefficient of O₂, ν is the kinematic viscosity of the electrolyte, and k is the electron-transfer rate constant.⁶⁸ According to eqn (3) and (4), the number of electrons transferred (n) during the ORR was calculated to be ∼3.0 per O₂ molecule at 0.2–0.5 V by using the values of C_O = 1.2 × 10⁻³ mol L⁻¹, D_O = 1.9 × 10⁻⁵ cm² S⁻¹ and ν = 0.01 cm² S⁻¹ in 0.5 M H₂SO₄ before the addition of PW, while n was obtained to be ∼4.0 at 0.2–0.5 V after addition of PWA. Furthermore, from the intercept of the Koutecky–Levich plots, kinetic limiting current density (J_K) was calculated to be 35.7 mA cm⁻² at 0.45 V for MWCNT/PVP/PWA/Pt that is more than that of MWCNT/PVP/Pt (9.5 mA cm⁻²) indicating the effect of PWA on electrocatalytic activity (Table 2). The rotating-disk voltammograms before and after PWA for ORR at the same rotating speed of 1000 rpm to represent the intrinsic activity of the catalysts were shown in Fig. 8(e). MWCNT/PVP/PWA/Pt exhibits a positive onset potential shift of 0.66 V than 0.62 V for MWCNT/PVP/Pt. The mass activity of MWCNT/PVP/PWA/Pt at the given potential (22.4 mA mg⁻¹ at 0.45 V) is 3.8 times higher than of MWCNT/PVP/Pt (5.86 mA mg⁻¹ at 0.45 V) measured under the similar conditions as presented in Fig. 8(f). MWCNT/PVP/PWA/Pt exhibited obviously enhanced specific activity (0.032 mA cm⁻² at 0.45 V), that is 3.6 also times that of the MWCNT/PVP/Pt (0.01 mA cm⁻² at 0.45 V) (Fig. 8(f)). The prominent ORR activity of the MWCNT/PVP/PWA/Pt compared to MWCNT/PVP/Pt implies that the absorbed PWA can affect the proton conductivity of the network structure of the obtained catalyst, thereby enhancing the ORR catalytic activity.

Table 2 Comparison of the number of electrons transferred, onset potential, J_k, rate constant and mass activity at 0.45 V of MWCNT/PVP/Pt and MWCNT/PVP/PWA/Pt catalysts, all Pt catalysts are in the same loading (1.59 mg cm⁻²) on the glassy-carbon electrode during tests

MWCNT/Pt	MWCNT/PVP/Pt	MWCNT/PVP/PWA/Pt
The number of electrons transferred	2.9	3.9
Onset potential [V]	0.62	0.66
Jk at 0.45 V (mA cm^−2)	9.33	35.7
Rate constant (cm s^−1)	2.74 × 10^−2	8.40 × 10^−2

---

Conclusions

We developed a composite, MWCNT/PVP/PWA/Pt, and introduced cross-linking between MWCNTs using PWA/PVP composite in order to use as a catalyst for fuel cell application. PVP in the MWCNT/PVP/PWA/Pt composite acts as a Pt adsorbing material, multiwall carbon nanotube (MWCNT)-solubilizing material, PWA immobilizer due to electrostatic interaction and proton conductor. Presence of PVP on MWCNTs surface makes them to repel each other in water and the results suggest that in the presence of PVP, distribution of nanoparticles improves significantly. The highly efficient and homogeneous loading of Pt nanoparticles onto the MWCNTs was realized by the PVP wrapping on the MWCNTs. A red shift frequency at Raman spectroscopy showed a strong reaction between the pyrrolidone chains and Pt nanoparticles. According to XPS analysis, the binding energy of N peaks has a positive shift after addition of PWA, this shift guaranties the interaction between PVP and PWA, thus, it can be confirmed that the N atom in PVP interacts with the proton, which results PWA/PVP acts as cross linker between MWCNTs and the formation of monolithic network and protons of PWA molecules jump to N atoms of PVP, therefore, the proton pathway is also formed by adjacent PWA particles and the side chain N-heterocycle moieties of PVP. The ORR measurements revealed that MWCNT/PVP/PWA/Pt nanocatalysts were more efficient than MWCNT/PVP/Pt and had better electrocatalytic activity for oxygen reduction. The thermal stability, high conductivity of the MWCNT/PVP/PWA/Pt hybrid nanocatalyst and lower activation energy, suggests a new nanostructured catalyst with the great potential application as a novel high-temperature (HT) polymer electrolyte membrane (PEM) fuel cell.

Acknowledgements

The authors gratefully acknowledge financial support from the Vice-Chancellor for Research Affairs of Tarbiat Modares University Dr Yaghoub Fathollahi and are grateful to Dr M. M. Hasani-sadrabadi, Dr Mohammad Momenian, Dr Fatemeh Molaabasi, Dr Leyla Irannezhad, Dr Ahmad Heydari, Dr Mehdi Hesani, Dr Mehdi Ghazanfari and Kourosh Rahimi for their scientific help and special thanks for Dr Saeede Ranjbari.

Notes and references

1. C. E. Thomas, Int. J. Hydrogen Energy, 2009, 34, 6005–6020.
2. S. Verhelst and T. Wallner, Prog. Energy Combust. Sci., 2009, 35, 490–527.
3. S. Giddey, F. T. Ciacchi, S. P. S. Badwal and V. Zelizko, Solid State Ionics, 2002, 153, 363–371.
4. M. Rikukawa and K. Sanui, Prog. Polym. Sci., 2000, 25, 1463–1502.
5. K. Ahmed and K. Föger, Ind. Eng. Chem. Res., 2010, 49, 7239–7256.
6. Y.-J. Wang, D. P. Wilkinson and J. Zhang, Chem. Rev., 2011, 111, 7625–7651.
7. A. Chandan, M. Hattenberger, A. El-kharouf, S. Du, A. Dhir, V. Self, B. G. Pollet, A. Ingram and W. Bujalski, J. Power Sources, 2013, 231, 264–278.
8. J. Zhang, Z. Xie, J. Zhang, Y. Tang, C. Song, T. Navessin, Z. Shi, D. Song, H. Wang, D. P. Wilkinson, Z.-S. Liu and S. Holdcroft, J. Power Sources, 2006, 160, 872–891.
9. Y. Shao, G. Yin and Y. Gao, J. Power Sources, 2007, 171, 558–566.
10. Y. Shao, J. Liu, Y. Wang and Y. Lin, J. Mater. Chem., 2009, 19, 46.
11. D. A. Stevens, M. T. Hicks, G. M. Haugen and J. R. Dahn, J. Electrochem. Soc., 2005, 152, A2309–A2315.
12. M. F. Mathias, R. Makharia, H. A. Gasteiger, J. J. Conley, T. J. Fuller, C. J. Gittleman, S. S. Kocha, D. P. Miller, C. K. Mittelsteadt, T. Xie, S. G. Yan and P. T. Yu, Electrochem. Soc. Interface, 2005, 14(3), 24–35.
13. J. Zhou, X. Zhou, X. Sun, R. Li, M. Murphy, Z. Ding, X. Sun and T. K. Sham, Chem. Phys. Lett., 2007, 437, 229–232.
14. S. Sharma and B. G. Pollet, J. Power Sources, 2012, 208, 96–119.
15. J. Rodriguez, Surf. Sci. Rep., 1996, 24, 223–287.
16. G. A. Attard and A. Bannister, J. Electroanal. Chem. Interfacial Electrochem., 1991, 300, 467–485.
17. J. Greeley, J. K. Nørskov and M. Mavrikakis, Annu. Rev. Phys. Chem., 2002, 53, 319–348.
18. B. Seger, A. Kongkanand, K. Vinodgopal and P. V. Kamat, J. Electroanal. Chem., 2008, 621, 198–204.
19. G. D. R. Cote, G. Lalande, G. Faubert, D. Guay and J. P. Dodelet, J. New Mater. Electrochem. Syst., 1998, 1, 7–16.
20. J. A. Schaidle, N. M. Schweitzer, O. T. Ajenifujah and L. T. Thompson, J. Catal., 2012, 289, 210–217.
21. I. Eswaramoorthi and A. K. Dalai, Int. J. Hydrogen Energy, 2009, 34, 2580–2590.
22. Z. Kónya, V. F. Puntes, I. Kiricsi, J. Zhu, J. W. Ager, M. K. Ko, H. Frei, P. Alivisatos and G. A. Somorjai, Chem. Mater., 2003, 15, 1242–1248.
23. K. Bouzek, K. M. Mangold and K. Jüttner, J. Appl. Electrochem., 2001, 31, 501–507.
24. H. Laborde, J. M. Léger and C. Lamy, J. Appl. Electrochem., 1994, 24, 1019–1027.
25. K. Jüttner, K. Bouzek and K.-M. Mangold, Electrochim. Acta, 2001, 46, 661–670.
26. J. Prakash, D. A. Tryk, W. Aldred and E. B. Yeager, J. Appl. Electrochem., 1999, 29, 1463–1469.
27. H. S. Oh, J. G. Oh, B. Roh, I. Hwang and H. Kim, Electrochem. Commun., 2011, 13, 879–881.
28. D. Villers, X. Jacques-Bédard and J.-P. Dodelet, J. Electrochem. Soc., 2004, 151, A1507.
29. P. Xu, X. Han, B. Zhang, Y. Du and H.-L. Wang, Chem. Soc. Rev., 2014, 43, 1349–1360.
30. M. J. O. Connell, P. Boul, L. M. Ericson, C. Hu, Y. Wang, E. Haroz, C. Kuper, J. Tour, K. D. Ausman and R. E. Smalley, Chem. Phys. Lett., 2001, 342, 265–271.
31. M. Okamoto, T. Fujigaya and N. Nakashima, Adv. Funct. Mater., 2008, 18, 1776–1782.
32. K. Matsumoto, T. Fujigaya, K. Sasaki and N. Nakashima, J. Mater. Chem., 2011, 21, 1187.
33. M. Shigeta, M. Komatsu and N. Nakashima, Chem. Phys. Lett., 2006, 418, 115–118.
34. M. Okamoto, T. Fujigaya and N. Nakashima, Small, 2009, 5, 735–740.
35. T. Fujigaya, M. Okamoto and N. Nakashima, Carbon, 2009, 47, 3227–3232.
36. T. Fujigaya, T. Uchinoumi, K. Kaneko and N. Nakashima, Chem. Commun., 2011, 47, 6843–6845.
37. M. M. Hasani-sadrabadi and E. Dashtimoghadam, Nanoscale, 2013, 11710–11717.
38. B. Wang, J. Power Sources, 2005, 152, 1–15.
39. K. C. Neyerlin, H. a. Gasteiger, C. K. Mittelsteadt, J. Jorne and W. Gu, J. Electrochem. Soc., 2005, 152, A1073.
40. G. O. Ilevbare and J. R. Scully, J. Electrochem. Soc., 2001, 148, B196.
41. H. Xu, Y. Song, H. R. Kunz and J. M. Fenton, J. Electrochem. Soc., 2005, 152, A1828.
42. M. Sun, H. Liu, Y. Liu, J. Qu and J. Li, Nanoscale, 2015, 7, 1250–1269.
43. H. J. Cui, H. M. Yu, J. F. Zheng, Z. J. Wang, Y. Y. Zhu, S. P. Jia, J. Jia and Z. P. Zhu, Nanoscale, 2016, 19–23.
44. O. Nakamura, I. Ogino and T. Kodama, Mater. Res. Bull., 1980, 15, 1049–1054.
45. I. V. Kozhevnikov and K. I. Matveev, Appl. Catal., 1983, 5, 135–150.
46. V. A. N. Giordano, A. S. Aricò, S. Hocevar, P. Staiti and P. L. Antonucci, Electrochim. Acta, 1993, 38, 1733–1741.
47. N. Mizuno and M. Misono, Chem. Rev., 1998, 2665, 199–217.
48. Y. Seung, F. Wang, M. Hickner, T. a. Zawodzinski and J. E. Mcgrath, J. Membr. Sci., 2002, 212, 263–282.
49. J. L. Malers, M. Sweikart, J. L. Horan, J. A. Turner and A. M. Herring, J. Power Sources, 2007, 172, 83–88.
50. R. Wlodarczyk, R. Wlodarczyk, M. Chojak and K. Miecznikowski, J. Power Sources, 2005, 159, 802–809.
51. J. F. Keggin, Proc. R. Soc. London, Ser. A, 1934, 75–100.
52. Y. Zhou, J. Yang, H. Su, J. Zeng, S. P. Jiang and W. a. Goddard, J. Am. Chem. Soc., 2014, 4954–4964.
53. O. Nakamura, T. Kodama, I. Ogino and Y. Miyake, Chem. Lett., 1979, 8, 17–18.
54. C. W. Lin, R. Thangamuthu and C. J. Yang, J. Membr. Sci., 2005, 253, 23–31.
55. S. S. Madaeni, S. Amirinejad and M. Amirinejad, J. Membr. Sci., 2011, 380, 132–137.
56. H. Chen, J. D. Snyder and Y. a Elabd, Macromolecules, 2008, 41, 128–135.
57. A. Saccà, A. Carbone, R. Pedicini, M. Marrony, R. Barrera, M. Elomaa and E. Passalacqua, Fuel Cells, 2008, 8, 225–235.
58. J. Zeng and S. P. Jiang, J. Phys. Chem. C, 2011, 115, 11854–11863.
59. J. Yuan, X. Jin, N. Li, J. Chen, J. Miao, Q. Zhang, L. Niu and J. Song, Electrochim. Acta, 2011, 56, 10069–10076.
60. A. Chen and P. Holt-hindle, J. Am. Chem. Soc., 2010, 3767–3804.
61. W. Li, C. Liang, W. Zhou, J. Qiu, Zhenhua, G. Sun and Q. Xin, J. Phys. Chem. B, 2003, 107, 6292–6299.
62. Y. Borodko, S. E. Habas, M. Koebel, P. D. Yang, H. Frei and G. a. Somorjai, J. Phys. Chem. B, 2006, 110, 23052–23059.
63. J. Zhang, H. Zou, Q. Qing, Y. Yang, Q. Li, Z. Liu, X. Guo and Z. Du, J. Phys. Chem. B, 2003, 107, 3712–3718.
64. A. Pozio, M. De Francesco, A. Cemmi, F. Cardellini and L. Giorgi, J. Power Sources, 2002, 105, 13–19.
65. M. J. Janik, R. J. Davis and M. Neurock, J. Am. Chem. Soc., 2005, 127, 5238–5245.
66. S. Lu, X. Xu, J. Zhang, S. Peng, D. Liang, H. Wang and Y. Xiang, Adv. Energy Mater., 2014, 4, 1–5.
67. J. J. Wang, G. P. Yin, J. Zhang, Z. B. Wang and Y. Z. Gao, Electrochim. Acta, 2007, 52, 7042–7050.
68. L. Guo, W.-J. Jiang, Y. Zhang, J. Hu, Z. Wei and L.-J. Wan, ACS Catal., 2015, 2903–2909.

---

Footnote

† Electronic supplementary information (ESI) available: Fig. S1–S3. See DOI: 10.1039/c6ra03509d

---