Self-assembled platinum nanoparticles on sulfonic acid-grafted graphene as effective electrocatalysts for methanol oxidation in direct methanol fuel cells

In this article, sulfonic acid-grafted reduced graphene oxide (S-rGO) were synthesized using a one-pot method under mild conditions, and used as Pt catalyst supports to prepare Pt/S-rGO electrocatalysts through a self-assembly route. The structure, morphologies and physicochemical properties of S-rGO were examined in detail by techniques such as atomic force microscope (AFM), transmission electron microscopy (TEM) and X-ray photoelectron spectroscopy (XPS). The S-rGO nanosheets show excellent solubility and stability in water and the average particle size of Pt nanoparticles supported on S-rGO is ~3.8 nm with symmetrical and uniform distribution. The electrocatalytic properties of Pt/S-rGO were investigated for methanol oxidation reaction (MOR) in direct methanol fuel cells (DMFCs). In comparison to Pt supported on high surface area Vulcan XC-72 carbon (Pt/VC) and Pt/rGO, the Pt/S-rGO electrocatalyst exhibits a much higher electrocatalytic activity, faster reaction kinetics and a better stability. The results indicate that Pt/S-rGO is a promising and effective electrocatalyst for MOR of DMFCs.


Results and Discussion
It is well known that GO contains a large amount of oxygen-containing groups such as hydroxyl, epoxy, carboxyl and carbonyl groups, which open up a versatile avenue for a variety of chemical transformations 39 . Among these groups, epoxy groups on the basal planes of GO can easily react with amines through a ring-opening reaction. In this work, 2-aminoethanesulfonic acid (taurine) was used to functionalize the GO. The process was carried out in aqueous solutions without adding additional organic solvents or surfactants that may generate undesirable byproducts or residues attaching on the surfaces of GO 7,40 . The zeta potential (ζ) values of the VC, GO, rGO and S-rGO suspensions were measured and the results are listed in Table 1. The zeta potential of VC is − 24.8 mV, close to − 27.6 mV measured on GO. The S-rGO aqueous suspension at a concentration around 0.01 mg mL −1 under neutral conditions has a zeta potential of − 62.4 mV, which is much lower than − 10.4 mV observed on rGO, indicating the success of the fuctionalization by taurine and the very good stability of the suspension 12,41 . Moreover, there is no sign of coagulation of S-rGO sheets after storing in water for 3 months (Fig. S1), indicating the high dispensability of S-rGO in water. Owing to the existence of negatively charged sulfonic acid group on the surface of S-rGO, self-assembly of the added positively charged polycation, poly(diallyldimethylammonium chloride) (PDDA) occurs as the result of the electrostatic interaction 42 . Thus, the charge inversion takes place on the S-rGO with the self-assembly of the positively charged PDDA monolayer, which can function as anchoring sites for negative charges or ligands. This leads to the self-assembly of the negatively charged Pt precursors, PtCl 6 2− , forming an uniformly distributed Pt precursors on the surfaces of S-rGO. After reduction treatment in NaBH 4 solution, uniformly distributed Pt NPs on the S-rGO supports were obtained. The one-pot synthesis route for S-rGO and Pt/S-rGO is shown in Fig. 1.
The morphologies of the GO and S-rGO were examined in details by AFM and TEM. The AFM images confirm that the GO and S-rGO are comprised of isolated and well dispersed GO sheets, as shown in Fig. 2. The GO sheets have lateral dimensions of several micrometers and a thickness of 0.8 nm, which is characteristic of a fully exfoliated GO sheet 12 . The thickness of a single-layer S-rGO sheet is about 1.2 nm. This implies that the taurine groups have been grafted onto GO, giving rise to the thickness increase. The single-layer sheets can also be observed clearly for both GO and S-rGO by TEM (Fig. 3). The corresponding selected-area electron diffraction (SAED) pattern of S-rGO coincides well with the typical SAED pattern of a single layer of rGO (inset in Fig. 3b) 43 . Figure 4 shows the XPS spectra of GO, rGO and S-rGO. As shown in the XPS survey scan of S-rGO, the peaks related to S2p at 168.5 eV and N1s at 399.8 eV can be clearly observed in the spectra of S-rGO (Fig. 4a). The deconvolution for C peaks was carried out by using an XPSPEAK software. A Shirley type background was used and the s peak type was selected. Adjustable parameters for the fitting were peak position, full width at half maximum, peak area and % Lorentzian-Gaussian. Four types of carbons were identified: C-O-C (286.7 eV), C-OH (286.0 eV), C= O (287.7 eV) and O= C-O (289.5 eV) in the spectra of the GO (Fig. 4f), which are in good agreement with the previous reports 13,44 . After the reduction, the peak intensities from oxygen-related groups decrease significantly (Fig. 4d,e). The peaks at 286.5 eV as shown in Fig. 4d can be attributed to the C-N groups.
The results indicate that taurine molecules have been successfully grafted to the GO, consistent with the AFM examinations (Fig. 2).   two peaks at 1178 cm −1 and 1560 cm −1 , indicating that the GO has been effectively reduced by NaBH 4 in mild conditions. For S-rGO, the absorbance bands at 1044 cm −1 and 1196 cm −1 can be attributed to the symmetric and asymmetric stretching band of -SO 3 groups 45,46 , which evidently demonstrate the successful grafting of taurine molecules on the GO and are in good accord with the AFM, XPS and zeta potential results. Raman spectroscopy  for the graphitic modes of carbon provided additional evidence for the structural changes caused by the reactions (Fig. S3). The samples VC, GO, rGO and S-rGO all exhibit D band at around 1586 cm −1 , which corresponds to a splitting of the E 2g stretching mode and reflects the structural intensity of the sp 2 -hbridized carbon atoms, and G band at around 1332 cm −1 . However, S-rGO shows slightly higher D/G intensity ratio than that of rGO, implying that more small domains of sp 2 structure were created, which is consistent with previous reports on functionalized graphene 5,47 .
Elemental analysis data (Table 1) indicate that the C, O and H contents in the VC, GO and rGO are within the range reported in literatures 5,47,48 . The N and S content in S-rGO is 2.3 and 5.2 wt.% of the total mass of S-rGO, respectively, and is evidently due to the grafted taurine molecules. This result indicates that taurine molecules exist in the as-prepared S-rGO composites. The electrical conductivities (σ, S cm −1 ) of the VC, GO, rGO and S-rGO films were measured by four-point probe method and are also listed in Table 1. The pristine GO shows a conductivity of ~2.4 × 10 −4 S cm −1 . The electrical conductivity of S-rGO is 4.3 S cm −1 , exhibiting a four-order magnitude increase in comparison with that of GO, although it is slightly lower than 5.2 S cm −1 of rGO. The results further verify that the π -conjugated system has been effectively restored in S-rGO after the reduction 49 . Figure 6 shows the TEM micrographs and the histograms of Pt NPs for Pt/VC and Pt/S-rGO electrocatalysts. The Pt loading is 20wt%. In the case of Pt/VC, the distribution of Pt NPs on VC is reasonable with slight aggregation and the average particle size of Pt NPs is ~4 nm. For the Pt/S-rGO, the average particle size is ~3.8 nm with symmetrical and uniform distribution. This indicates that S-rGO is an effective supports for Pt NPs. Fig. 7 is the TEM micrographs of 40 wt.% Pt/S-rGO and 80 wt.% Pt/S-rGO electrocatalysts. As can be seen, the average particle size increases slightly with increasing the metal loading. However, the size dispersion is still very uniform even in the case of high Pt loading of 80 wt.%. These results clearly indicate that the sequential functionalization and self-assembly process is very effective for the synthesis of uniform and nano-sized Pt/S-rGO electrocatalysts with high loading owing to the existence of extensive negatively charged sulfonic acid groups on the surface of rGO.
The electrocatalytic activity of the Pt catalysts on the different supports for MOR was examined at 25 °C in a nitrogen-saturated 0.5 M H 2 SO 4 solution in the absence and presence of 1.0 M CH 3 OH solution and the results are shown in Fig. 8. The electrochemical surface area (ECSA) was obtained from the hydrogen adsorption/ desorption area in the CV curves measured in a 0.5 M H 2 SO 4 solution (Fig. 8a). The higher ECSA in general represents the more active sites for the methanol oxidation. The Pt/S-rGO electrocatalyst shows an ECSA with a value of 394.2 cm 2 mg −1 Pt , much higher than Pt/VC (286.6 cm 2 mg −1 Pt ) and Pt/rGO (292.4 cm 2 mg −1 Pt ). Considering the similar sizes of the Pt NPs supported on S-rGO and VC supports (see Fig. 6), the significantly high ECSA of the Pt/S-rGO may indicate the synergestic effect between Pt NPs and S-rGO supports. As shown in Fig. 8b, the mass specific peak current density for MOR is 206, 263 and 465 mA mg −1 for Pt/VC, Pt/rGO and Pt/S-rGO, respectively. The much higher mass specific catalytic activity of the Pt/S-rGO again indicates that S-rGO plays a significant role to promote the electrocatalytic activity of Pt NPs. Meanwhile, the sulfonic acid groups would be helpful to form hydroxyl groups from water activation, which is essential for removing the metal-poisoning CO molecules from the surfaces of catalysts 37,50 . This appears to be supported by the high I f / I b (the forward peak current density/the backward peak current density) ratio of Pt/S-rGO (1.0), as compared to I f /I b ratio of Pt/VC (0.65) and Pt/rGO (0.89), as the peak current I f /I b ratio could be used to estimate the CO tolerance on Pt surface.
The onset potentials are 0.63, 0.60 and 0.57 V for the reaction on Pt/VC, Pt/rGO and Pt/S-rGO, respectively. The negative shift of onset potential indicates an improvement in the electrocatalytic reaction kinetics 51,52 . Tafel curves (potential vs. log i) were plotted to compare the MOR kinetics of Pt/VC, Pt/rGO and Pt/S-rGO. As shown in Fig. 8c, the Tafel plots can be divided into two nearly linear regions which intersect at approximately 0.55 V for the reaction, indicating a change in the rate determining steps. In the region below 0.55 V, the Tafel slopes of MOR on Pt/VC and Pt/rGO are 122 and 118 mV dec −1 , very close to the theoretical value predicted for one-electron transfer reaction as rate determining step. However, for Pt/S-rGO, the Tafel slope in the same region is much higher, 156 mV dec −1 , indicating a faster methanol dehydrogenation even at the relatively low overpotential region 53 .  The stability of the electrocatalysts was evaluated by a chronoamperometry (CA) method at a constant potential 0.6 V vs. SCE (Fig. 8d). As can be seen, polarization current drops significantly with testing time, which is very rapid at the initial stage, followed by a much slower decay. The rapid current decay represents the poisoning effect on the catalyst surface. However, the Pt/S-rGO shows the slowest current decrease among the electrocatalysts studied. After polarized at 0.6 V for 1800 s, the polarization current for the MOR on Pt/S-rGO electrocatalysts is 61 mA mg −1 Pt , significantly higher than 22 mA mg −1 Pt on Pt/VC and 20 mA mg −1 Pt on Pt/rGO. This indicates a better tolerance of Pt/S-rGO electrocatalysts to CO or carbonaceous intermediates formed during the MOR.
The effect of the grafted sulfonic acid groups on the electronic structure of Pt was further studied by XPS and the results are shown in Fig. 9. The Pt4f spectra of the samples can be fitted with the following components: Pt 0 4f 7/2 , Pt 0 4f 5/2 , Pt 2+ 4f 7/2 , Pt 2+ 4f 5/2 , Pt 4+ 4f 7/2 and Pt 4+ 4f 5/2 . The Pt 0 4f 7/2 and Pt 0 4f 5/2 peaks of Pt/S-rGO show a negative shift as compared to that on Pt/VC and Pt/rGO. For example, the binding energy for Pt 0 4f 7/2 on Pt/S-rGO is 71.7 eV, lower than 71.9 eV observed on Pt/VC and Pt/r-GO. The negative shift of the d-band center of Pt is most likely originated to the grafted sulfonic acid group which is electronegative and may donate electrons to Pt. According to the d-band center theory, the CO adsorption would be weakened as the d-band center decreases 54 . Thus, the negative shift of d-band center of Pt NPs on S-rGO presents a weaker oxophilicity, facilitating the removal of the adsorbed poisoning intermediates such as CO ads for MOR 55 . This is supported by the significantly higher I f /I b ratio and high current density after polarization at 0.6 V for 1800 s (Fig. 8b,d). The oxidation state of Pt NPs was analyzed and calculated based on the XPS Pt4f spectra, and the results are given in Table 2. The Pt/S-rGO electrocatalyst shows the highest metallic contents of Pt (i.e., Pt(0)) that is important for MOR 56 . The taurine groups assembled on the rGO frameworks significantly enhances the electrocatalytic activity of Pt NPs and facilitate the transfer of electrons and protons for the MOR.
The charge transfer capabilities of the different electrocatalysts were further investigated by electrochemical impedance spectroscopy (EIS) in a 10 mM K 3 [Fe(CN) 6 ] + 0.01 M PBS solution. In the Nyquist plots as shown in Fig. 10, the semicircular portion at high frequencies corresponds to the charge transfer limited process and the linear portion at low frequency region can be ascribed to the diffusion process. The impedance data were fitted using an equivalent circuit model with NOVA 1.9 software (Metrohm, Switzerland). The electrode charge transfer resistance, R ct was obtained. The decreased diameter of the semicircle reflects the decrease in the R ct . Based on the equivalent circuit, the R ct values were determined to be 352, 203, 185 and 158 Ω on the bare GCE, Pt/VC Pt/rGO and Pt/S-rGO electrodes, respectively. These results clearly confirm that R ct of Pt/S-rGO is significantly lower as compared to that on Pt/VC and Pt/rGO, most likely due to the good ionic conduction ability of sulfonic acid groups grafted on the rGO.
Finally, the synthesis method of the current work was compared to relevant reports on the functionalization of graphene and proposed enhancement mechanism for Pt electrocatalysts (see Table 3). There are similar processes to the current work, but the present study is the first time to report taurine functionalized graphene as Pt   catalyst support for MOR and to clarify in details the effects of grafted sulfonic acid groups for the promotion of the dispensability of graphene and catalytic effect on the supported Pt NPs. The existence of sulfonic acid groups on graphene not only enhances the dispensability and stability of graphene, and most important, it modifies the electronic structure and negatively shifts the d-band center of Pt, leading to the reduction in charge transfer resistance and enhanced activity of the Pt electrocatalysts for MOR.

Conclusion
A facile and simple one-pot method was developed to graft taurine molecules onto the basal planes of rGO, and subsequently used as effective Pt catalyst support for MOR via a self-assembly route. The synthesis process was carried out in aqueous solutions without adding any other organic solvents or surfactants. The self-assembled Pt/S-rGO electrocatalyst shows a significantly higher electrocatalytic activity, faster reaction kinetics and a much better stability for MOR in comparison with conventional Pt/VC and Pt/rGO catalysts. The catalytic enhancements could be attributed to the following factors: (1) the presence of large number of sulfuric acid functional group on rGO due to taurine functionalization; (2) uniform deposition of Pt NPs and smaller particle size as result of the self-assembly process; (3) the negative shifted d-band center and increased charge transfer ability due to the grafted sulfonic acid groups. The results in this report indicate that it is feasible to functionalize graphene through the newly developed one-pot method and show the promising potential to enhance the practical applications of graphene as effective catalyst support for MOR in DMFCs.

Methods
Synthesis of S-rGO. GO was prepared from the natural graphite by a modified Hummers method 57 . The details are provided in the supplementary information. To synthesize S-rGO, 600 mg GO was dispersed into 200 mL deionized (DI) water under ultrasonication. The suspension was centrifuged at 4000 rpm for 30 min to remove the small amount of unexfoliated particles. The obtained supernatant was transferred into a 500 ml flask purged by N 2 , followed by adding 12 g taurine (≥ 99.5%, Sigma-Aldrich). The mixture was then heated at 60 °C for 36 h under stirring. After cooling to room temperature, a 200 mL with 0.25M sodium borohydride (NaBH 4 , Sigma-Aldrich) aqueous solution was added to the mixture under stirring for 4 h. The mixture was then vacuum filtered through 0.2 μ m polycarbonate membrane and washed with DI water to remove the residual ions and excessive raw materials. The as-prepared S-rGO was freeze-dried for 48 h and stored in a dry cabinet for further use.   electrocatalysts with different Pt loading (20, 40 and 80 wt.%) were prepared. The 20 wt.% Pt/VC carbon black were synthesized using the same procedures as described above. Pt NPs supported on rGO were prepared by a chemical reduction method using NaBH 4 as the reducing agent. Firstly, GO (20 mg) was dispersed in 100 mL DI water by ultrasonication for 30 min before 5 mL H 2 PtCl 6 solution (10 mg/mL) was added to the solution. Excess NaBH 4 with a weight of 3 g was slowly added to the solution with magnetic stirring for 24 h at room temperature. The resulting hybrids were separated by filtration and washed with DI water for several times. The Pt loading was controlled at ~20 wt.%.

Synthesis of Pt
Characterization. AFM measurements were performed in the tapping mode using Digital Instruments Nanoscope DI 3100. The morphology of the catalysts was studied by a transmission electron microscope (JEOL 2010) operating at 200 kV. FTIR measurements were carried out with a Perkin Elmer Model GX spectrometer.
Raman spectra of the different electrocatalysts were obtained by using a Renishaw ® RM1000 Raman spectroscope. XPS measurements were performed on a Kratos Analytical AXIS His spectrometer with a monochromatized Al Ka X-ray source (1486.6 eV photons), in a vacuum of 10 −8 Torr at a constant dwelling time of 100 ms and pass energy of 40 eV. The anode voltage was 15 mV with a current of 10 mA. All the survey scans and the core-level spectra were obtained at photo-electron take-off angel of 90° (α , with respect to the sample surface). Survey scans were recorded within a range of 0-1200 eV. To compensate for the effects of surface charging, all core-level spectra were referenced to the C1s hydrocarbon peak at 284.5 eV. Elemental analysis was performed using Perkin Elmer Instruments CHNS-O Analyzer. The electrical conductivity of the samples was measured on a resistivity/sheet resistance measurement station (Advanced Instrument Technology, Model: CMT-SR2000N) using a four probe method. The films were prepared by vacuum filtration onto alumina membranes (Millipore, 0.02 μ m pore size, 47 mm membrane diameter) supported on a fritted glass holder. After washing with excess DI water, the films were dried in a vacuum oven at 40 °C for 24 h. The thickness of the films was controlled at ~20 μ m as measured by using an Alpha-step 500 profilometer. The conductivity was calculated using the formula: σ = 1/ (R s t), where R s is the resistance (unit: Ω ) and t is the film thickness (unit: cm). Zeta potential (ζ) was measured on a Zetasizer Nano Series 90 at 25 °C, using a DTS 1060 C disposable capillary cell (Malvern Instruments, UK).
Electrochemical measurements. The electrochemical measurements were carried out in a conventional three-electrode cell using a potentiostat/galvanostat (Autolab PGSTAT30) at ambient temperature. A glassy carbon electrode (GCE, 4 mm in diameter) was used as the working electrode, a platinum wire as the counter electrode, and a saturated calomel electrode (SCE, 0.241 V versus RHE) as the reference electrode. A salt bridge with a Luggin capillary was connected to the reference electrode. To load the electrocatalyst suspension onto GCE, the electrocatalyst powder was ultrasonically mixed in isopropanol to form a homogeneous ink with the catalyst concentration of 5 mg mL −1 , followed by dropping 10 μ L of the electrocatalyst ink onto the surface of GCE. Then, 10 μ L 0.5 wt.% Nafion solution was coated onto the surface to fix the electrocatalysts. Without specification, all the electrochemical potentials reported in the present study were given versus RHE. The ECSA of Pt/S-rGO and Pt/VC electrocatalysts was measured in a nitrogen-saturated 0.