Abstract

Physicochemical and electrochemical characterisations of Pt-based electrocatalysts supported on carbon (Vulcan carbon, C1, and carbon produced by plasma pyrolysis of natural gas, C2) toward ethanol electrooxidation were investigated. The Pt20/C180 and Pt20/C280 electrocatalysts were prepared by thermal decomposition of polymeric precursors at 350°C. The electrochemical and physicochemical characterisations of the electrocatalysts were performed by means of X-ray diffraction (XRD), transmission electron microscope (TEM), Raman scattering, cyclic voltammetry, and chronoamperometry tests. The XRD results show that the Pt-based electrocatalysts present platinum metallic which is face-centered cubic structure. The results indicate that the Pt20/C180 electrocatalyst has a smaller particle size (10.1–6.9 nm) compared with the Pt20/C280 electrocatalyst; however, the Pt20/C280 particle sizes are similar (12.8–10.4 nm) and almost independent of the reflection planes, which suggests that the Pt crystallites grow with a radial shape. Raman results reveal that both Vulcan carbon and plasma carbon are graphite-like materials consisting mostly of sp2 carbon. Cyclic voltammetry and chronoamperometry data obtained in this study indicate that the deposition of Pt on plasma carbon increases its electrocatalytic activity toward ethanol oxidation reaction.

1. Introduction

Global warming and methods to avoid it have been the topics of much recent debate. Fear of exhausting fossil fuel-based energy sources, the price and supply crises for this type of fuel, and clean and renewable energy policies have motivated alternative energy sources. Such concerns support implementing of the fuel cells, which were discovered by Sir William Grove in 1839 and are a form of efficient electricity generation with low environmental impact [1, 2].

Ethanol has been used as a fuel and is much less toxic, compared with methanol; and the technology involved in the manufacture of the ethanol fuel cell is similar to that of direct methanol cells [3]. In Brazil, ethanol production is especially efficient; the industry maintains such production through a well-structured methodology due to the use of ethanol as vehicle fuel over several decades [4].

A support material commonly used in electrocatalyst for fuel cells is the Vulcan carbon. This type of carbon is widely employed in electrocatalyst preparation because it has a surface area that facilitates maximum nanoparticle dispersion and adequate electrocatalyst pore distribution. Moreover, the pore size and functional groups on the carbon surface are also critical for its use as an electrocatalyst support [5]. Recently, different kinds of carbon nanomaterials such as hollow graphitic nanoparticles [6], carbon nanotubes [7], and graphitic carbon nanofibers [6] were investigated and different fabrication methods have been adopted. These carbon materials can exhibit superior performance compared to the conventional carbon supports for EOR (ethanol oxidation reaction) anode catalysts. Carbon materials usually have a high surface area to disperse metal grains, and their high conductivity transfers electrons generated from electrochemical reactions taking place on the anode [8].

Tang et al. [9] reviewed new carbon materials such as ordered porous carbon and carbon nanofibers, which have been used in direct alcohol fuel cells. According to them, these materials generally presented better performance due to their special structure, better crystallinity, good stability, and faster mass transfer compared to the commercial carbon nanotubes materials [10]. Moreover, it has been shown that carbon can also be produced by plasma pyrolysis of natural gas [11]. This method is a promising way of producing high-purity black carbon without generation of environmentally harmful products [11]. Upon plasma pyrolysis, methane decomposes to produce hydrogen and a solid carbon rich residue commonly designed as “black plasma” [12, 13].

The Pt-based electrocatalysts are the most used on Polymer Membrane Fuel Cells (PMFC) for oxidation of hydrogen [14, 15] as well as low molecular weight organic molecules such as methanol [16, 17], ethanol [18, 19], ethylene glycol [20, 21], and dimethyl ether [22, 23]. These fuels have been widely investigated not only in fundamental research but also in several commercial corporations [24]. However, the high cost and the scarcity of Pt form a great border in PMFC’s development. A fundamentally interesting feature of these Pt-based electrocatalysts is their size and morphology dependent electrocatalytic activity [25]. It is well known that pure Pt can be readily poisoned by CO-like intermediate species produced in the alcohol oxidation reaction [26, 27]. Therefore, it is very important to investigate new low loading of Pt materials that not only can improve the CO-tolerance of Pt, but also enhance its catalytic activity for alcohol oxidation.

Thus, the aim of this paper was to prepare Pt-based electrocatalysts on Vulcan carbon, C1, and carbon produced by plasma pyrolysis of natural gas, C2; then, the electrocatalyst activity toward ethanol electrooxidation was compared.

2. Experimental Section

2.1. Synthesis of Pt/C Electrocatalysts by Thermal Decomposition of Polymeric Precursors (DPP)

Thermal decomposition of polymeric precursors (DPP) was used to prepare Pt/C electrocatalysts with a nominal composition, Pt20/C180 and Pt20/C280 (where C1 is the Vulcan carbon, purchased from Carbot Corporation and C2 is a carbon produced by plasma pyrolysis of natural gas, which was provided by Prof. Dr. Alfredo Gonçalves Cunha).

The Pt polymeric precursor resin, Pt-resin, was synthesised by mixing citric acid (CA) in ethylene glycol (EG) in the proportion 1 : 4 at 60–65°C. After the citric acid was fully dissolved in ethylene glycol, 50 mL of a solution of metal precursor salt (H2PtCl6, Aldrich) dissolved in isopropanol (~0.02 mol/L) was slowly added to the CA/EG mixture. After the solution was fully added, the H2PtCl6/CA/EG mixture temperature was raised to 80–85°C for esterification. The H2PtCl6/CA/EG mixture was kept under stirring for 0.5–1 h, which generated the desired Pt-resin. The Pt-resin produced using this method was stable and can be stored at room temperature; the resin metal concentration was determined using inductively coupled plasma optical emission spectrometry (ICP-OES) with the Pt-resin = 3.8 × 10−4 mol Pt/g resin.

The Pt/C electrocatalysts dispersed on carbon (C1, Vulcan carbon, or C2, plasma carbon) with the mass ratio 20 wt.% platinum and 80 wt.% carbon were prepared by adding the respective Pt-resin precursors at levels corresponding to the desired compositions (Pt20/C180 and Pt20/C280). Two millilitres of ethanol was added to the mixture to aid dispersion in an ultrasonic bath for 30 minutes. Both the Pt20/C180 and Pt20/C280 electrocatalysts were annealed at 350°C in an air atmosphere for 3 hours using an oven.

2.2. Physicochemical Characterisations of Pt/C Electrocatalysts

Physicochemical characterisation included X-ray diffraction (XRD) using a Bruker D8 advance with CuKα radiation ( Å) and scanning from 20° to 90° (step = 0.03° and step time = 3 s). The crystallite sizes () were determined using the Scherrer equation (1), as follows [28]:where is the reflection width at half-maximum intensity and is the angle at the maximum intensity.

The two different carbon and Pt/C electrocatalysts were analysed through Raman spectroscopy using a Witec Confocal Raman Microscope System Alpha 300R. The Raman analyses were performed over the material using a neon laser incidence at 533 nm and the wavenumbers range 0–4000 cm−1.

The morphology and dispersion of Pt/C electrocatalyst particles were investigated using a JEOL/JEM-1400 transmission electron microscope (TEM) operated at 200 kV.

2.3. Electrochemical Characterisations of Pt/C Electrocatalysts

Electrochemical characterisations were performed using 0.5 mol dm−3 H2SO4 (Aldrich) as the supporting electrolyte in ultra-pure water (18 MΩ cm at 20°C). The graphite electrode (4 cm2 geometric area) and reference electrode were used as the counter and reference electrodes, respectively.

The working electrode was prepared through deposition on a carbon substrate previously polished using alumina (0.3 μm) and 30 μL of a Pt/C electrocatalyst solution, which was prepared by dispersing 1 mg of electrocatalyst in ethanol (190 μL) and Nafion (10 μL). The materials electrocatalytic activities were evaluated through cyclic voltammetry (CV), which was recorded in the −100 mV to 1000 mV versus potential range with and without 1.0 mol dm−3 ethanol (Aldrich) and chronoamperometry at 400 mV versus for 2 hours. The electrochemical experiments were performed at room temperature in a one-compartment electrochemical cell (body, 50 mL) using an AUTOLAB 302N model potentiostat/galvanostat.

3. Results and Discussion

Figure 1 shows the XRD patterns for the Pt20/C180 and Pt20/C280 electrocatalysts, which were prepared by thermally decomposing the polymeric precursors. The carbon support (peak at = 25°) was detected. The XRD patterns show the primary peaks for a face-centred cubic (FCC) crystalline Pt (PDF: 01-087-0646 [29]) with (111), (200), (220), and (311) diffraction planes. The crystallite sizes were measured using Scherrer’s equation [28], and the results are shown in Table 1. The results indicate that the Pt20/C180 electrocatalyst has a smaller particle size (10.1–6.9 nm) compared with the Pt20/C280 electrocatalyst. However, the Pt20/C280 particle sizes are similar (12.8–10.4 nm) and almost independent of the diffraction planes, which suggests that the Pt crystallites grow with a radial shape.

The TEM images for both Pt20/C180 and Pt20/C280 electrocatalysts are shown in Figure 2. One can see that the particles are not well dispersed on C1 support. The higher particle density and particle size in Pt20/C280 electrocatalyst strongly suggest that nucleation is promoted in this sample. Because this effect occurs in the Vulcan carbon support, one can infer that surface defects in this material act as enhanced nucleation sites for Pt DPP process. However, the electrocatalyst Pt20/C280 is well dispersed and its particle size (average ca. 6.1 nm) agrees with XRD data. It is obvious from these results that the particle size of Pt increases when the Vulcan carbon is used as material support compared with the plasma carbon, which leads to some undesirable agglomeration of the particles.

Figure 3 shows the Raman spectra of carbon materials and Pt/C electrocatalyst prepared by DPP process. The well-known peaks of carbon graphite [3032], which are the primary features, were observed for all materials investigated with varying levels of intensity and width in the 1350 and 1590 cm−1 regions, respectively. Moreover, in the carbon plasma sample (Figure 3(c)), the other strong band near 2680 cm−1, corresponding to a 2D mode, is a relatively sharp peak that can be associated with a graphene layer [33]. However, in the Pt20/C280 sample, this strong band disappears; it might be due to Pt nanoparticles presence or to a thermal treatment during DPP process. The intensity of G peak (~1580 cm−1) and ratio (2D)/(G) in Figures 3(a) and 3(c), respectively, reveals that both Vulcan carbon and plasma carbon are graphite-like materials consisting mostly of sp2 carbon [34].

Figure 4(a) shows the cyclic voltammograms (CV) for the two Pt/C electrocatalysts deposited on a carbon substrate in 0.5 mol dm−3 H2SO4. The CV curves are distorted compared with that of pure Pt (data not shown here). Moreover, the hydrogen oxidation region (from −0.10 to 0.20 V versus ) for both Pt20/C180 and Pt20/C280 electrocatalysts does not show the typical adsorption/desorption peaks, conversely to the case of Pt/C with 40% metal loading published by us before [35]. The obtained CV profile, suggesting a partially blocked Pt surface, may be due to the incomplete decomposition of carbonaceous species from DPP process. The Pt20/C280 has a larger double layer region compared with Pt20/C180 electrocatalysts, which may be associated with surface higher area in the plasma carbon than in the Vulcan carbon, because it is well known that carbon is responsible for large capacitive responses.

Cyclic voltammograms for oxidation of the ethanol adsorbed on the Pt/C electrocatalysts are compared in Figure 4(b). The cyclic voltammetry curves for ethanol show three peaks: one during the positive-going scan (Vulcan carbon  V versus ;  mA cm−2, and plasma carbon  V versus ;  mA cm−2) which is attributed to the ethanol oxidation and two oxidation peaks observed during the negative-going scan which is associated with the removal/or oxidation of intermediates species (e.g., CO, , and ) not completely oxidized in the positive-going scan [21]. The ethanol begins to oxidise at ~0.4 V versus , and the electrocatalyst fuel oxidation efficiency can be analysed; ethanol oxidation is more efficient where the oxidation potential is lower and the current density is higher (demonstrated using the CV) [35]. For example, if one fixes the potential at 0.5 V versus one gets the following current density values: Vulcan carbon:  mA cm−2 and  mA cm−2; plasma carbon:  mA cm−2 and  mA cm−2. Thus, one can infer that Pt20/C280 displays better performance than Pt20/C180 electrocatalysts and this behavior indicates a possible structure effect in the Pt electrocatalytic activity. The structure effect in the Pt electrocatalytic activity in the EOR has been demonstrated by other studies on well-ordered [36, 37] as well as nanostructured surfaces [38, 39].

Figure 5 represents the chronoamperometry (CA) curves recorded at a constant potential of 400 mV versus for two hours. CA allowed evaluation of the electrocatalysts activity of the electrocatalysts. This is a typical exponential decay observed in any step-potential technique; however, one can observe that, during the first minutes, there is a sharp decrease in the current density normalized by Pt loading for the Pt20/C180 electrocatalyst, followed by a slow decrease in the -values at 2500 s, which become very lower after this period time. This behavior could be explained by strong adsorption of the intermediates CO, , and type residues on the Pt active sites, which is causing the poisoning of the surface. It is well known in the literature that and CO species could poison the platinum sites during the alcohol electrooxidation reaction [26, 27, 40]. However, for the Pt20/C280 electrocatalyst, the behavior is very different; one can see that current density normalized by Pt loading decreases slowly for longer time periods. In this case, the main cause of slow current-time decay is not due to poisoning of the platinum sites but is due to the surface instability of nanoparticles, because of the surface recrystallisation, surface metal segregation, particle agglomeration, and so forth, decreasing slowly the number of active sites of the electrocatalysts [35]. The Pt20/C280 electrocatalyst furnished the best result in this test, because to the end of the experiment the -values obtained for the Pt20/C280 (5 mA/gPt) are higher than for the Pt20/C180 electrocatalyst.

4. Conclusions

This paper has shown that Pt-based electrocatalysts can be prepared using carbon produced using a plasma torch from natural gas vehicles. XRD results indicate that the Pt20/C180 electrocatalyst has a smaller particle size (10.1–6.9 nm) compared with the Pt20/C280 electrocatalyst (12.8–10.4 nm). Moreover XRD data obtained for Pt20/C280 suggests that the Pt crystallites grow with a radial shape. Raman results showed that all the materials are composed of randomly oriented structural units varying from 50 to 80 Å in size. The voltammetry for the Pt20/C280 electrocatalyst showed better charge density and current, compared with the Pt20/C180 electrocatalyst. Finally, CA experiment has demonstrated that Pt20/C280 electrocatalyst furnished the best result, thus demonstrating its enhanced oxidative capacity for ethanol.

Conflict of Interests

The authors declare that there is no conflict of interests regarding the publication of this paper.

Acknowledgments

J. Ribeiro especially acknowledges FAPES and CNPq for financial support. The authors also acknowledge CAPES, FINEP, LMC, and PETROBRAS.