Nanocomposite for methanol oxidation: synthesis and characterization of cubic Pt nanoparticles on graphene sheets.

We present our recent results on Pt nanoparticles on graphene sheets (Pt-NPs/G), a nanocomposite prepared with microwave assistance in ionic liquid 2-hydroxyethanaminiumformate. Preparation of Pt-NPs/G was achieved without the addition of extra reductant such as hydrazine or ethylene glycol. The Pt nanoparticles on graphene have a cubic-like shape (about 60 wt% Pt loading, Pt-NPs/G) and the particle size is 6 ± 3 nm from transmission electron microscopy results. Electrochemical cyclic voltammetry studies in 0.5 M aqueous H2SO4 were performed using Pt-NPs/G and separately, for comparison, using a commercially available electrocatalyst (60 wt% Pt loading, Pt/C). The electrochemical surface ratio of Pt-NPs/G to Pt/C is 0.745. The results of a methanol oxidation reaction (MOR) in 0.5 M aqueous H2SO4 + 1.0 M methanol for the two samples are presented. The MOR results show that the ratios of the current density of oxidation (If) to the current density of reduction (Ib) are 3.49 (Pt-NPs/G) and 1.37 (Pt/C), respectively, with a preference by 2.55 times favoring Pt-NPs/G. That is, the tolerance CO poisoning of Pt-NPs/G is better than that of commercial Pt/C.

Introduction

Synthesis methods of nanocomposites for platinum on n class="Chemical">graphene have fascinated physicists, chemists and material scientists since the Nobel Laureates of physics, Geim and Novoselov, found the single-layer graphene using tape exfoliation in 2004 [1]. Graphene is one of the two-dimensional materials for the 21st century. However, because of unstable thermodynamic properties, graphene had been predicted to exist for a long time but was not actually produced until recently [2-8]. Graphene appears as a unique sp2-bonded carbon network in single-layered structure. Possible applications of graphene are numerous. Energy is one of the fields where graphene can be utilized [9, 10]. Graphene or carbon nanotubes (CNT) may serve as a support material to anchor nanoparticle catalysts in order to improve electron transport, which provides possibilities in the design of next-generation catalysts with enhanced interactions between the substrate and the catalyst [11, 12]. Since the chemical exfoliation method can also oxidize the surface of graphene, part of the graphene structure is actually graphene oxide.

Noble metal nanoparticles have been decorated on n class="Chemical">carbon materials for use in direct methanol fuel cell (DMFC) applications ever since nanotechnology was in its developing stage. The carbon materials for DMFC or proton exchange membrane fuel cell application usually possess high specific surface area, electrical conductivity, chemical stability and so on [11-13]. The commonly used carbon materials for DMFC are active carbon, Vulcan XC-72 carbon black, multi-walled/single-walled CNT and graphene (such as graphene oxide or functionalized graphene) [14-16]. In order to improve the solubility of carbon materials in fuel cell applications, sulfonation was applied to CNT, carbon black or active carbon for better solubility prior to making the devices. In this study, the graphene layers were also sulfonated to give better water affinity [17, 18].

Since the expensive Pt is the best transition n class="Chemical">metal for fuel cell technology, its electrocatalytical efficiency should be balanced with its cost in fabrication. The well-dispersed nanoparticles, Pt or Pt–M alloys (M for Ru, etc), on electrodes are the desirable conditions in practice [19-22]. A lower Pt loading with a higher electrocatalytical property is the optimum goal. Herein the synthesis of graphene materials follows the procedures outlined below: (i) strong oxidation of graphite to produce graphite oxide (GO), (ii) surface modification of GO with simultaneous reduction of GO to graphene and (iii) reduction reaction for the precursors of noble metal ions (and simultaneously for GO) to become nanoparticles on graphene sheets. The cubic Pt nanoparticles were usually synthesized with reducing agents and surfactants, such as citrate and oleic acid or polyvinylpyrrolidone [23-26]. One possible barrier for the Pt electrode in the DMFC application is its CO poisoning. CO is constantly present during the methanol oxidation reaction in that CO likely occupies the active sites on the Pt electrode, drastically reducing the activity [27-29].

Ionic liquids, as green chemicals for use in environmentally friendly synthesis, have enhanced physical properties suitable for replacing common organic solvents, especially the ones used at high temperatures. The advantages of ionic liquids are (a) almost no vapor pressure, (b) a much greater temperature rann class="Chemical">ge as a liquid and (c) varying solubility for organics, inorganics and polymers. Ionic liquids have found numerous uses as reaction and extraction media [30-32].

The 2-hydroxyethanaminium formate was introduced by Bicak et al in 2005 [33]. Later, Mudring et al synthesized porous n class="Chemical">silver materials using 2-hydroxyethanaminium formate with microwave assistance [34]. Without an extra reducing agent, the Ag+ was reduced by the ionic liquid itself. With liberation of carbon dioxide, the formate salt was known to produce hydrogen that reduces the Ag+ into Ag0. Taking this precedent example as a model synthesis, we have succeeded in the preparation of cubic Pt nanoparticles on graphene sheets (Pt-NPs/G) using this simple 2-hydroxyethanaminium formate as a reactive solvent with microwave assistance.

Herein, we describe the synthesis of cubic Pt-n class="Chemical">NPs/G nanocomposites for methanol oxidation reaction (MOR) application to compare with a commercial one having a similar electrocatalyst weight loading of about 60 wt%. The green chemical synthesis method employed the ionic liquid that was acting as the medium and as the reducing agent. The resistance to CO-poisoning of the electrocatalytical nanocomposites reported in this study was better than that of commercial ones as evidenced in the MOR faradic current density ratios (vide infra).

Experimental

Modified graphene sheets

The graphene synthesis followed a modified Hummer's method [31]. Concentrated n class="Chemical">sulfuric acid (25 ml) was poured into a 250 ml round-bottomed flask held in an ice bath. Then fumed nitric acid (10 ml) was added slowly in 15 min. Graphite powder (1 g) was introduced under vigorous stirring into the flask kept in the ice bath. After being mixed well, the solution was added with potassium chlorate (22 g) in 30 min. The ice bath was removed and the solution was stirred at room temperature for 96 h. Suitable amounts of deionized (DI) water were added to the product mixture kept in the ice bath and the solution was centrifuged for removal of the liquid phase. The procedure was performed three times. The solids collected were rinsed with methanol three times. The mud-like solid was dried at 80 °C for 12 h to yield the GO.

Sulfonation of GO was carried out using the following procedure: 250 mg n class="Chemical">graphene oxide in 250 ml DI water was stirred for 30 min before the addition of 1.4 g NaBH4, and kept at 80 °C for 1 h. After centrifugation, the mud-like residuals were rinsed with methanol three times and then dried at 80 °C in an N2 atmosphere for 1 h. After being dried, the mixture of the above residuals (158 mg) in 300 ml DI water was dispersed in a round-bottomed flask for 30 min using an ultrasonic bath. Sulfanilic acid (140 mg) together with potassium nitrate (50 mg) was introduced into a 100 ml beaker containing DI water (40 ml), employing an ice bath. After being mixed well, 1 N HCl (1 ml) was added to the solution and then the solution was poured into the round-bottomed flask and stirred for 2 h in the ice bath. Centrifugation followed by removal of aqueous solution resulted in the sulfonated graphene, which was rinsed with methanol a few times and then dried at 80 °C in an N2 atmosphere.

The microwave-assisted synthesis of the Pt-n class="Chemical">NPs/G composite was performed using a CEM Discover Du7046 microwave set with 20 W power output for 30 s to increase the temperature to 80 °C and then held at 80 °C for 5 min. Three samples were prepared with graphene or sulfonated graphene (100 mg) as substrates and with grinded K2PtCl6 at 355, 355 or 100 mg, respectively, plus 2-hydroxyethanaminium formate (5 g), in Pyrex glass tubes. The cubic Pt-NPs/G was the one with the K2PtCl6/sulfonated graphene w/w ratio of 3.55. The preparation of 2-hydroxyethanaminium formate (shown in scheme 1) was done by a slow neutralization of H2NCH2CH2OH (20 ml) and formic acid (14 ml) in a 100 ml round-bottomed flask kept in the ice bath.

Scheme 1.

Preparation of 2-hydroxyethanaminium formate.

XRD characterization

Cubic Pt-n class="Chemical">NPs/G crystalline materials mixed with estimated 1 wt% graphite were examined using the PANalytical X'Pert Pro MPD powder x-ray diffractometer (XRD) with Cu Kα radiation at 50 kV and 40 mA. The scanning rate was 0.625° min−1 and the 2θ range was 5–80°.

Fourier transform infrared (FT-IR) characterization

The samples for FT-IR analysis consisted of composite materials (5 mg) with dried KBr (100 mg), well mixed and ground and then pressed into pellets for measurement. The Perkin Elmer spectrum 100 FT-IR was scanned from 400 to 4000 cm−1, signals being accumulated 32 times and resolution about 0.01 cm−1.

Pt content measurement with thermal gravimetric analysis and inductively coupled plasma mass spectrometry (ICP-MS)

Thermal gravimetric analysis (TGA) was performed using the Perkin Elmer Pyris 1 TGA (the resolution was about 0.1 μg) with about 5 mg sample loading on the platinum plate, from room temperature to 900 °C at a heating rate of 5 °C min−1. The Pt content was measured alternatively using an Agilent 7700 Series ICP-MS.

Transmission electron microscopy (TEM) characterization

The TEM analysis was performed on a JEOL 2010FX with a LaB6 electron gun operated at 200 kV. The TEM samples were prepared with ethanol dispersion dropped on the grid Cu net for analysis. The selected area electron diffraction (n class="Chemical">SAED) was employed to reveal the cubic Pt-NPs/G electronic diffraction pattern and the lattice distances (d-spacing).

X-ray photoelectron spectroscopic (XPS) analysis

A Physical Electronics PHI Quantra XPS microprobe equipped with an Al Kα monochromatic x-ray source (1486.6 eV, 45° incident angle) was employed on the cubic Pt-n class="Chemical">NPs/G and the commercial Pt/C samples.

Electrochemical analysis

Cyclic voltammetry (CV) was employed with the conventional three-electrode glass cell Radiometer Voltalab 40 at room temperature. The glass-carbon working electrode was pre-polished, and then n class="Chemical">alcohol immersed. The working electrode was surface covered with 20 μl dispersions of cubic Pt-NPs/G (13.75 μg cm2) or alternatively that of Pt/C (17.50 μg cm2), which was the commercial 60 wt% Pt (Uni Ward Co. Ltd, Taiwan). The reference electrode was Ag/AgCl and the counterelectrode was a Pt wire. The electrochemical surface area (ECSA) was evaluated in 0.5 M aqueous H2SO4 and the scanned voltage from −0.2 to 1.2 V with a scan rate at 20 mV s−1. The methanol oxidation was measured at 20 mV s−1 scan rate from 0 to 1.0 V in 0.5 M aqueous H2SO4 + 1 M aqueous methanol.

Results and discussions

Synthesis and characterization of cubic Pt-NPs/G

The results of chemical exfoliated reaction are shown in figure 1 with the FT-IR spectra for graphene (n class="Chemical">GE), sulfonated graphite oxide (GO-SO3H) and graphene oxide (GO). The assignments of the peaks on GO were 1715 cm−1 for and 1580 cm−1 for vibration modes, 1260–1210 cm−1 for the C–O–C mode and 1053 cm−1 for C–O bonds. The assignments of peaks on GO-SO3H were 3500–3750 cm−1 peaks for N–H bonds, 2359 cm−1 for nitrile, 1696 cm−1 for and 1526 cm−1 for and 1263 and 1106 cm−1 for –SO3H bonds. There were weak peaks for C–O–C and C–O bonds around 1100–1300 cm−1. The peak assignments on GE were 1750 cm−1 for bonds, and weak peaks at 1200–1000 cm−1 for the C–O bonding mode. The broad peaks around 3000–3500 cm−1 were assigned as –OH bonds with the order of broadening GO > GO-SO3H > GE.

Figure 1.

FT-IR spectra of GE, GO-SO3H and GO.

FT-IR spectra of GE, n class="Chemical">GO-SO3H and GO.

The three cubic Pt-n class="Chemical">NPs/G composites were different in w/w ratios of sulfonated graphene/K2PtCl6 or GO-SO3H/K2PtCl6 at 1/3.35, 1/3.35 and 1/1 in the preparation. Only the ones with the w/w ratio of sulfonated graphene/K2PtCl6 at 1/3.35 yielded the cubic Pt-NPs/G composite with particle size at 6 ± 3 nm and Pt loading was estimated to be at 62.83 wt% by the TGA examination (shown in figure 3(a)) and at about 61.82 wt% by the ICP-MS measurement. The same ratio at 1/3.35 for GO-SO3H/K2PtCl6 produced cubic Pt-NPs/GO with about 40.08 wt% Pt loading and an average particle size of 18 ± 5 nm. The surface oxygenated functional groups had affected apparently the nucleation on nanoparticle synthesis. If the ratio was lowered to 1/1 for GO-SO3H/K2PtCl6, the Pt-NPs/GO formed exhibited a smaller average particle size. Figure 2 shows the TEM results and the particle size distribution of Pt nanoparticles on cubic Pt-NPs/G composite. Table 1 reveals the results of sulfonated graphene/K2PtCl6 and graphene oxide/K2PtCl6 in different w/w ratios. There was no extra reducing agent employed during the synthesis, the ionic liquid 2-hydroxyethanaminium formate also playing the role of a reducing agent.

Figure 3.

TGA results for estimating Pt contents of the three synthesized samples and (b) XRD diffractogram of cubic Pt-NPs/G with 1 wt% graphite added.

Figure 2.

TEM characterization of cubic Pt-NPs/G composite: (a) at 15 000 magnification, (b) a closer look, (c) Pt nanoparticle size distribution and (d) SAED analysis of the image.

Table 1.

Synthesis of Pt-NPs/G with 2-hydroxyethanaminium formate ionic liquid with microwave assistance.

	Loading (mg)
K2PtCl6	Ionic liquid	Substrateb (100 mg)	Shape/size (nm)
1	355	15 000	GE	Cubic/6±3
2	355	15 000	GO	Cubic/18±5
3	100	15 000	GO	Spheric/14±4

20 W power output for 30 s and then the reaction vessel was kept at 80 °C for 5 min.

Sulfonated substrates.

Synthesis of Pt-n class="Chemical">NPs/G with 2-hydroxyethanaminium formate ionic liquid with microwave assistance.

The ionic liquid 2-hydroxyethanaminium formate was utilized by Mudring et al in the synthesis of the n class="Chemical">monolith silver, a porous material, together with the microwave heating technique without the addition of an extra reducing agent. Ag0 was reduced from Ag+. Mudring had observed the formation of CO2 gas in the synthesis with the above-mentioned ionic liquid. We believe that the current formation of Pt0 nanoparticles proceeds in a similar way. The Pt4+ metal ions were reduced to Pt0 at the expense of the formate that was concurrently oxidized to CO2. Scheme 2 shows the possible redox pathway, parallel to the findings of Mudring and co-workers [34].

Scheme 2.

The proposed reduction mechanism of 2-hydroxyethanaminium formate.

The XRD results (shown in figure 3(b)) are consistent with the fact that the graphene sheets anchor the n class="Chemical">Pt nanoparticles with (111), (200) and (220) crystalline planes indexed with an fcc structure at 2θ = 39.5°, 46.3° and 67.8°, respectively. The XRD diffractogram was recorded on the cubic Pt-NPs/G and graphite mixture in the w/w ratio of 100:1 to demonstrate that the graphene and the graphite crystalline peaks were at different 2θ values of 25° and 28.4°. The graphene XRD peak was broad and the graphite peak was sharp, because the graphite is with the stacked layers of graphene but not the graphene as a result partly with the oxidation and sulfonation reactions in preparation. The current synthetic method successfully produced graphene sheets decorated with Pt nanoparticles.

TGA results for estimating Pt contents of the three synthesized samples and (b) XRD diffractogram of cubic n class="Chemical">Pt-NPs/G with 1 wt% graphite added.

The amount of Pt nanoparticles on n class="Chemical">Pt-NPs/G composite was measured with TGA. The Pt loadings on samples in table 1, entries 1–3, were estimated to be about 62.83, 40.08 and 12.78 wt%, respectively. The ICP-MS result indicated that the Pt loading of cubic Pt-NPs/G was about 61.83%. As shown in figure 2 the TEM analysis gives images of the cubic Pt-NPs/G started initially from sulfonated graphene to Pt precursors in a w/w ratio of 1/3.55. As shown in figure 2(d), the SAED reveals that the cubic Pt-NPs/G has a platinum fcc (111) crystalline lattice with a d-spacing of 0.21 nm. In this study, the cubic Pt-NPs/G nanocomposite was chosen to be one of the candidates for MOR, taking advantage of the smallest size of Pt nanoparticles among samples and the cubic shape. For performance comparison towards the methanol oxidation, a commercial Pt/C electrocatalyst was selected with specifications of 60 wt% Pt loading and particle size distribution in the range 3–5 nm.

The XPS analysis was performed to look in more detail at the Pt chemical environments and n class="Chemical">carbon materials carrying surface oxygenated functional groups. The results suggest that the cubic Pt-NPs/G and Pt/C have very similar environments. Figure 4 reveals the two peaks assigned to Pt 4f7/2 and 4f5/2 for cubic Pt-NPs/G and Pt/C composites and both resolved into the Pt0, Pt2+ and Pt4+ components. The two composites indeed have almost the same relative component percentages of Pt.

Figure 4.

XPS spectra of cubic NPs/G and Pt/C composites: (a) Pt 4f XPS of cubic Pt-NPs/G, (b) Pt 4f XPS of Pt/C, (c) C 1s XPS of cubic Pt-NPs/G, (d) C 1s XPS of Pt/C, (e) O 1s XPS for Pt-NPs/G and (f) O 1s XPS for Pt/C.

XPS spectra of cubic NPs/G and Pt/C composites: (a) n class="Chemical">Pt 4f XPS of cubic Pt-NPs/G, (b) Pt 4f XPS of Pt/C, (c) C 1s XPS of cubic Pt-NPs/G, (d) C 1s XPS of Pt/C, (e) O 1s XPS for Pt-NPs/G and (f) O 1s XPS for Pt/C.

The results might imply that the Pt nanoparticles produced with the ionic liquid n class="Chemical">2-hydroxyethanaminium formate with microwave assistance are not different from the Pt nanoparticles produced commercially. The Pt0 species in cubic Pt-NPs/G and that in Pt/C are 44.4 and 45.56%, respectively. The metallic state Pt0 is more associated with methanol oxidation and electrochemical activity [35, 36].

XPS data provide only the surface property and not the bulk property, because the depth of analysis is about 10 nm. There are more surface oxyn class="Chemical">genated functional groups on the graphene prepared in this study than those of the carbon black from commercial sources. The results indicate the presence of hydroxyl and epoxy groups. The sp2-C–C percentage of the cubic Pt-NPs/G is estimated to be 42.42% and that of the commercial Pt/C 48.02%. The carbon black is generally made by the high-temperature sintering method so that the electrical conductivity is suitable for fuel cell catalyst support in terms of enhanced catalytic activity and corrosion resistance. The corrosion behavior might affect the performance and stability of the Pt catalyst [37]. The oxygenated functional groups on the surface are further divided into hydroxyl and epoxy groups (from C1s XPS narrow band spectrum data analysis), totaling 57.58% on the cubic Pt-NPs/G (C–OH group for 27.29%, and C–O–C group for 30.29%) and 51.93% on Pt/C (C–OH 23.27% and C–O–C 28.66%). O 1s XPS spectra were shown in figures 4(e) and (f). Both cubic Pt-NPs/G and Pt/C spectra were fitted to the three peaks as phenol (about 384 eV), C–OH (about 382–383 eV) and C = O (about 381 eV). The C–OH ratio of cubic Pt-NPs/G was larger than the Pt/C, suggesting in part more anti-CO-poisoning of the cubic Pt-NPs/G (see figure 4 and table 2).

Table 2.

XPS peak position and percentage contributions of cubic Pt-NPs/G and Pt/C.

	Pt 4f7/2	Pt4+	Pt 4f5/2	Pt4+	C 1s	C–O–C	O 1s	Phenol
Pt0	Pt2+	Pt0	Pt2+	sp2	C–OH		C–OH
Cubic Pt-Nps/G (eV)	70.6	71.3	72.6	74.3	75.3	75.9	284.0	285.1	286.2	531.2	532.2	534.8
%	24.4	11.7	9.2	20.0	18.9	15.9	42.4	27.3	30.3	14.2	55.3	30.5
Pt/C (eV)	70.7	71.5	72.4	74.2	75.2	76.1	283.8	285.2	286.1	529.8	531.7	533.6
%	23.3	12.6	6.4	22.2	16.4	19.1	48.7	23.3	28.7	19.9	48.3	31.8

ECSA analysis

The CV analysis for ECSA was measured in 0.5 M aqueous H2SO4 at 20 mV s−1 scan rate from −0.2 to 1.2 V versus the Ag/n class="Chemical">AgCl reference electrode. For comparison the CV results of two Pt-containing electrocatalysts were stacked in figure 5. In the range −0.2 to 0.1 V (versus Ag/AgCl) in potential, the marked areas (in black and red dashed lines) are the ECSA for the two samples, as calculated using the equation where QH (μC cm−2) presents the value of hydrogen adsorption of the Pt active sites on the surface. The constant 210 is the charge required to oxidize a monolayer of H2 on the single-crystal Pt surface. The Pt loading is presented in μg cm−2. The ECSA result for cubic Pt-NPs/G was 24.16 m2 g−1 and that of the commercial Pt/C was 32.45 m2 g−1. The ratio of Pt-NPs/G to commercial Pt/C was 0.745. The ECSA of commercial Pt/C is greater than that of cubic Pt-NPs/G.

Figure 5.

Cyclic voltammograms of cubic Pt-NPs/G on the working electrode and of Pt/C on the working electrode.

Cyclic voltammograms of cubic Pt-n class="Chemical">NPs/G on the working electrode and of Pt/C on the working electrode.

Methanol oxidation reaction

The MOR was measured in 0.5 M aqueous H2SO4 + 1.0 M aqueous n class="Chemical">CH3OH at a scan rate of 20 mV s−1, with electrocatalysts as the glass-carbon working electrode and a Pt wire as the counterelectrode. The loading of cubic Pt-NPs/G was 13.75 μg cm−2 and that of Pt/C was 17.50 μg cm−2.

The current densities, normalized by glass-carbon area, are shown in figure 6(a). The forward faradic anodic peak at about 0.6–0.9 V versus Ag/n class="Chemical">AgCl was identified as If (methanol oxidation) and the backward faradic scanning peak at about 0.2–0.7 V versus Ag/AgCl was identified as Ib (CO additive formation on the Pt electrocatalyst surface). The onset potentials of If were 0.28 and 0.20 V for cubic Pt-NPs/G and Pt/C, respectively. The forward anodic maximum peak potentials of methanol oxidation were assigned at 0.70 and 0.74 V versus Ag/AgCl for the cubic Pt-NPs/G and the Pt/C, respectively. The backward scan maximum peak potentials were 0.53 and 0.5 V versus Ag/AgCl for the cubic Pt-NPs/G and the Pt/C, respectively. The difference of methanol oxidation peak potentials was highly correlated with the oxygenated functional groups on the surfaces of the graphene sheets or of the carbon black. Overall, the If/Ib ratios implied the CO-poisoning attribution of the electrocatalysts [19, 27].

Figure 6.

MOR of cubic Pt-NPs/G and of Pt/C in 0.5 M aqueous H2SO4 + 1 M aqueous methanol, at 20 mV s−1 scan rate: (a) normalized by area and (b) normalized by Pt electrocatalyst loading and area.

MOR of cubic Pt-n class="Chemical">NPs/G and of Pt/C in 0.5 M aqueous H2SO4 + 1 M aqueous methanol, at 20 mV s−1 scan rate: (a) normalized by area and (b) normalized by Pt electrocatalyst loading and area.

Figure 6(a) shows that the methanol oxidation ability follows the ECSA results of the cubic n class="Chemical">Pt-NPs/G and of the Pt/C, i.e. that of Pt/C and that of cubic Pt-NPs/G. However, the If/Ib ratios were in reversed order: 3.49 for the cubic Pt-NPs/G and 1.37 for the Pt/C. The If of Pt/C to the If of cubic Pt-NPs/G was 2.39 times greater and the Ib of Pt/C to the Ib of cubic Pt-NPs/G was 6.1 times greater. Thus, the CO additive formation on the Pt/C modified working electrode ought to be much faster than that on the cubic Pt-NPs/G modified working electrode. It would be problematic in long-term application of methanol oxidation if the active sites of oxidation are occupied by CO.

The current densities, normalized by Pt loading weight, are shown in figure 6(b). The n class="Chemical">methanol oxidation CV suggests that the current density per mg of cubic Pt-NPs/G is better than that of Pt/C. The If ratio (cubic Pt-NPs/G to Pt/C) was 0.53 and the Ib ratio was 0.21 (cubic Pt-NPs/G to Pt/C). Based on ECSA results the ratio of electrocatalyst active sites between cubic Pt-NPs/G and Pt/C was 0.745. Based on MOR results the ratios of If and Ib were both lower than 0.745. However, figure 6(b) indicates that the If peak height of cubic Pt-NPs/G is greater than that of Pt/C; in other words, the cubic Pt-NPs/G has a higher CO tolerance.

Given in figure 7(a) is the larger aggregation morphology for the commercial n class="Chemical">Pt/C. Figures 7(b) and (c) are higher-magnification images for the aggregation morphology and SAED pattern of the same Pt/C. A lower magnification image is shown for comparison with the similar magnification image of cubic Pt-NPs/G in figure 2(a). The well-dispersed Pt particles on the graphene sheets were apparently more preferred than that of Pt/C. The SAED pattern exhibits a d-spacing of about 0.23 nm. The aggregation surely reduces the active areas for electrocatalysis and hence the catalytic activity decreases.

Figure 7.

TEM images of commercial Pt/C aggregation morphologies: (a) low-magnification image, (b) SAED pattern and (c) high-magnification image of (a).

TEM images of commercial n class="Chemical">Pt/C aggregation morphologies: (a) low-magnification image, (b) SAED pattern and (c) high-magnification image of (a).

With cubic Pt nanoparticles anchored on the surface of graphene sheets, the ECSA and MOR results suggest that the advantages of the cubic Pt-NPs/G composite may likely lie in the CO-antipoisoning aspect.

Conclusions

In this study, we produced the cubic Pt nanoparticle decorated n class="Chemical">graphene sheets (cubic Pt-NPs/G) in ionic liquid 2-hydroxyethanaminium formate with microwave assistance. The ratio of sulfonated graphene to the Pt precursor is 1/3.55 for better dispersion in aqueous solutions, resulting in the cubic Pt-NPs/G nanocomposite. There is neither hydrazine nor ethylene glycol serving as the added reductant to convert the Pt precursor to Pt nanoparticles. The Pt loading of cubic Pt-NPs/G is 60% by weight from the TGA results. The Pt crystallites of cubic Pt-NPs/G were in an fcc structure and the Pt particle sizes were about 5–10 from TEM and SAED results. The ECSA of cubic Pt-NPs/G is smaller than that of commercial Pt/C. So is the MOR ability. The ECSA for cubic Pt-NPs/G and that for Pt/C are 24.16 and 32.45 m2 g−1, respectively, whereas the MOR If/Ib ratio for cubic Pt-NPs/G and that for Pt/C are 3.49 and 1.37, respectively. A high MOR If/Ib ratio indicates that cubic Pt-NPs/G as the electrocatalyst is the better one for long-term fuel cell applications.