Direct methanol fuel cells (DMFCs) have received great attention as promising candidates for the renewable energy system in portable electronic devices and unmanned aerial vehicles [1-5]. Pt has been used as a typical catalyst for the methanol oxidation reaction (MOR) at the anode of the DMFCs [5-7]. However, Pt catalysts are vulnerable to poisoning by carbon monoxide (CO) produced during the electrochemical oxidation of methanol [8-10]. In addition, owing to the high cost and low abundance of Pt, the practical use of the DMFCs has been limited for decades. To address these issues, significant efforts have been made toward the development of bimetallic structures by alloying Pt with another transition metal, and thus overcome the CO poisoning effect, reduce the Pt loading, and improve the MOR activity [3, 11].

Until recently, Pt-Ru bimetallic alloy catalysts have been considered as the best CO tolerant materials [4, 12-15]. The addition of an oxophilic Ru to Pt can provide adsorbed hydroxyl species (OH_ads) near the Pt catalytic sites, which serve to oxidize the poisoning species (CO) to CO₂. In this mechanistic manner, Pb has recently emerged as a more oxophilic element than Ru, not only to promote the CO oxidation but also to optimize the electronic structure of Pt, resulting in superior electrocatalytic properties toward the MOR [16-19]. For instance, Disalvo and coworkers [19] reported that the Pt-Pb nanoparticles showed enhanced specific activity toward the MOR as compared to that of commercial Pt black and Pt-Ru nanoparticles.

The synthesis of Pt-based bimetallic nanocrystals with controllable shapes and structures has been extensively investigated in order to achieve high performance of the MOR catalyst [16, 20, 21]. The anisotropic morphology of Pt-based nanoparticles is more interesting than that of conventional nanoparticles for the MOR electrocatalysis. In particular, two-dimensional (2D) nanoplates have recently been demonstrated as a good model catalyst because of their large ratio of high-coordination surface atoms, which can be beneficial for the catalytic activity and long-term performance [22-24]. As an example, Huang and coworkers [21] reported that Pt-Pb/Pt core/shell nanoplates showed 2.4 and 7.9 times higher MOR activity than Pt-Pb nanoparticles and the commercial Pt/C, respectively.

Based on recent reports, we hypothesized that the MOR properties can be improved more effectively by modulating the surface atomic distribution of the morphology-controlled nanocrystals. Therefore, in this study, we controlled the Pt and Pb atomic distributions on the surface of the Pt-Pb bimetallic hexagonal nanoplates to investigate the surface elemental effect for the enhancement of the MOR. After a typical synthesis, a homogeneous distribution of the Pt-Pb alloy nanoplates was obtained. By selective etching of Pb through the acetic acid (HAc) treatment, the Pt skin and a part of the heterogeneous Pt-Pb alloy surface were observed. Electrochemical measurements revealed that the MOR specific and mass activity of the Pt-Pb hexagonal nanoplates were 1.3 and 1.1 times higher than those of the HAc-treated Pt-Pb nanoplates, respectively. After 1000 cycles of the durability test, the MOR mass activity of the Pt-Pb hexagonal nanoplates showed a 20.9% reduction, indicating better durability compared to that of the HAc-treated Pt-Pb nanoplates (45.0% reduction). In addition, after 3000 s of the chronoamperometry test, the current for the Pt-Pb hexagonal nanoplates was 1.4 times higher than that for the HAc-treated Pt-Pb nanoplates. In conclusion, the Pb distribution on the Pt-Pb alloy surface aided its high CO tolerance, superior activity, and stability during the MOR electrocatalysis.

Platinum(II) 2, 4-pentanedionate (Pt(acac)₂, 99%, Alfa Aesar), lead(II) 2, 4-pentanedionate (Pb(acac)₂, 99%, Alfa Aesar), trioctylphosphine (TOP, 97%, Sigma-Aldrich), benzyl ether (BE, 98% Sigma-Aldrich), oleylamine (OAm, 70%, Sigma-Aldrich), HAc (≥99.7%, Duksan), toluene (99.5%, Duksan), ethyl alcohol (99%, Duksan), Nafion perfluorinated ion-exchange resin (5 wt% solution, Sigma-Aldrich), isopropyl alcohol (IPA, > 99.5%, Duksan), methanol (CH₃OH, ≥99.9%, Sigma-Aldrich), and perchloric acid (HClO₄, 70%, Sigma-Aldrich) were used as received.

In a typical procedure, a vial containing Pt(acac)₂ (10 mg, 0.025 mmol), Pb(acac)₂ (8 mg, 0.024 mmol), OAm (4 mL), and BE (6 mL) was ultrasonicated for 5 min and heated at 190 ℃ for 2 min. Then, TOP (0.2 mL, 0.45 mmol) was added into the vial, after which the vial was kept at 190 ℃ for another 2 h under magnetic stirring. The resulting product was cooled to room temperature and separated from the solution by centrifugation at 3000 rpm for 10 min after adding toluene (10 mL). Then, the resulting Pt-Pb nanoplates were re-dispersed in toluene (10 mL) for further use.

A toluene solution (10 mL) containing Pt-Pb nanoplates was added into another toluene (10 mL) solution containing Vulcan XC-72 carbon black, and then, the mixture was ultra-sonicated for 2 h and washed with ethanol (10 mL) by centrifugation at 3000 rpm for 10 min. The final Pt-Pb nanoplates/C catalyst was dried in an oven at 70 ℃ for 30 min.

The Pt-Pb nanoplates/C catalyst was added to a concentrated HAc solution (20 mL) and heated at 60 ℃ for 30 min. The final Pt-Pb HAc/C catalyst was washed twice with ethanol and dried in an oven at 70 ℃ for 30 min.

Transmission electron microscopy (TEM) images were obtained using an HT-7100 microscope (Hitachi, Japan) operated at an acceleration voltage of 120 kV. Scanning electron microscopy (SEM) images were obtained using a Jeol JEM-7210F (Tokyo, Japan). High-resolution TEM was performed using a Titan G2 ChemiSTEM Cs Probe (FEI, Netherlands) operated at an acceleration voltage of 200 kV. X-ray diffraction (XRD) data were obtained with a D2 phaser X-ray diffractometer (Bruker, USA). The Pt and Pb loadings in the catalysts were analyzed by inductively coupled plasma-atomic emission spectroscopy (ICP-AES, PerkinElmer, Optima 7300DV).

The electrochemical characterization of the catalysts was performed at room temperature, utilizing a three-compartment electrochemical cell connected to a potentiostat (CHI 600E from CH Instruments and VSP from Bio-Logic). A Pt mesh (1 × 1 cm²) and Ag/AgCl electrodes were used as the counter and reference electrodes, respectively. The potentials were converted to a reversible hydrogen electrode (RHE). A glassy carbon disk (GC, 5 mm in diameter, PINE instrumentation) was used as the working electrode and polished with a 0.05 μm alumina suspension. The commercial Pt/C (20 wt%) was bought from Alfa Aesar.

The catalyst ink was prepared by adding a catalyst powder, 5 wt% Nafion (10 μL), and IPA (0.25 mL) to deionized water (1.0 mL) under ultrasonic agitation for 10 min. The ink was dropped on the GC electrode using a micropipette and dried to form a thin film. The total Pt loading on the GC electrode was 10.2 μg cm^–2 for each catalyst. The electrode was pre-cycled in an Ar-saturated (ultrahigh purity, 99.999%) 0.1 mol L^–1 HClO₄ solution for 100 cycles between 0.08 and 1.20 V, at a scan rate of 100 mV s^–1. Then, cyclic voltammograms (CVs) were recorded by cycling between 0.08 and 1.20 V, at 50 mV s^–1, in an Ar-saturated 0.1 mol L^–1 HClO₄ solution.

In a typical CO stripping experiment, the pre-cleaned electrode was held at a potential of 0.05 V for 5 min in a CO-saturated (ultrahigh purity, 99.95%) 0.1 mol L^–1 HClO₄ solution. The CO stripping curve was then recorded by cycling between 0.08 and 1.20 V at 50 mV s^–1 after the extra CO in the solution was removed by purging Ar for 10 min.

MOR was carried out with CV curve by sweeping the potential between 0.08 and 1.27 V, at a scan rate of 50 mV s^–1, in an Ar-saturated solution containing 0.1 mol L^–1 HClO₄, and 1.0 mol L^–1 CH₃OH. The long-term durability test of the catalyst was performed by applying a cyclic potential sweep between 0.6 and 1.1 V, at a scan rate of 100 mV s^–1 for 1000 cycles, in an Ar-saturated 0.1 mol L^–1 HClO₄ and 1.0 mol L^–1 CH₃OH solution at room temperature. Then, the CV was recorded again in a fresh Ar-saturated 0.1 mol L^–1 HClO₄ and 1.0 mol L^–1 CH₃OH solution and compared with the initial CV of the MOR. The chronoamperometry curve of the commercial Pt/C catalyst was recorded for 1000 s, and the curves for the Pt-Pb nanoplates/C and Pt-Pb HAc/C were recorded for 3000 s in an Ar-saturated 0.1 mol L^–1 HClO₄ solution and a 1.0 mol L^–1 CH₃OH solution, at 0.9 V.

In a typical synthesis of Pt-Pb hexagonal nanoplates, Pt(acac)₂ and Pb(acac)₂ were added to a mixture of OAm and TOP as co-surfactants and reducing rate controllers, and BE as the solvent, under Ar atmosphere. The mixture was heated at 190 ℃ for 2 h under magnetic stirring. The resulting product was cooled to room temperature and collected by centrifugation with toluene. Fig. 1a and Fig. S1 (see the Supporting Information) show the TEM images of the as-obtained Pt-Pb hexagonal nanoplates with an average edge length of 43.3 ± 4.5 nm. The SEM image of the Pt-Pb nanoplates revealed a plate-like morphology with an average thickness of 10.02 ± 1.1 nm (Fig. S2) [25]. The high-resolution TEM and scanning TEM-high-angle annular dark field (STEM-HAADF) images of a single Pt-Pb nanoplate are shown in Figs. 1b and c. Elemental mapping based on energy-dispersive X-ray spectroscopy (EDS) confirmed that Pt and Pb were distributed homogeneously on the hexagonal plate. The enlarged high-resolution TEM images of the edge, vertex, and inner areas of the Pt-Pb nanoplate are shown in Figs. 1d and e. In the side and inner parts, lattice spacings of 0.213 and 0.366 nm, respectively, indicated a typical Pt-Pb alloy with (110) and (100) planes, respectively [21, 26]. In addition, the entire area of the plate showed a single crystalline structure, demonstrating the formation of a homogeneous Pt-Pb alloy. The fast Fourier transform (FFT) pattern obtained from the high-resolution TEM image projected along the [001] zone axis is shown in Fig. 1f [21]. The ordered hexagonal-like spot arrays illustrated in the FFT pattern reconfirmed the formation of a single crystalline Pt-Pb bimetallic alloy structure.

To investigate the evolution of the Pt-Pb hexagonal nanoplates, a set of control experiments were conducted by altering the reaction parameters. Fig. S3 shows the TEM images of the products obtained in the typical reaction conditions, except that different volumes of OAm were used. Upon adding half of the initial volume of OAm (2 mL), nanoplates with a shorter edge length and irregular-shaped nanocrystals were observed (Fig. S3a), whereas increasing the volume of OAm to 6 mL generated larger hexagonal nanoplates (Fig. S3b). Based on these observations, we demonstrated that OAm acts not only as a surfactant but also as a coordination ligand. It has been reported that the amine group of OAm can coordinate with Pt ions to form a metal complex, resulting in slower reduction of the Pt ion to metal in the initial stage of the synthesis [27-29]. Therefore, in this work, the formation of a smaller number of seeds in the presence of a larger volume of OAm promotes the size growth of the Pt-Pb nanoplates. Fig. S4 shows the TEM images of the products obtained in another control experiment where the volume of TOP was altered. In the absence of TOP, Pt-Pb agglomerates were generated as the major product. Increasing the volume of TOP from 0.1 to 0.3 mL led to a change in shape from rod-like nanocrystals to nanoplates. There were no products in the presence of 0.4 mL of TOP during the synthesis. These results revealed that TOP, in the appropriate volume, acts not only as a co-surfactant but also as a controller of the reduction kinetics during the preparation of the Pt-Pb nanoplates [30, 31].

The Pt-Pb nanoplates were loaded on carbon supports (Fig. S5, Pt-Pb nanoplates/C) for further characterization and treatment without particle agglomeration. To control the surface composition of the Pt-Pb nanoplates, we treated the sample with HAc at 60 ℃ for 30 min [11]. The crystal structure of the Pt-Pb nanoplates/C before and after the HAc treatment was characterized by XRD. The XRD reflection peaks for both the samples were typical and in good agreement with the index of the JCPDS reference for the Pt-Pb bimetallic alloy crystal structure (Fig. S6). To gain more structural insight, TEM and high-resolution TEM images of the HAc-treated Pt-Pb nanoplates/C (Pt-Pb HAc/C) were recorded (Figs. 2a and b, respectively). The enlarged high-resolution TEM image of a Pt-Pb HAc nanoplate shows lattice spacings of 0.228 nm, indicating Pt(111) at the edge part and 0.214 nm of Pt-Pb(110) in the core part (Fig. 2c) [21]. The FFT pattern of the selected area of the nanoplate, labeled by a white box in Fig. 2c, is shown in Fig. 2d, along the [001] zone axis. The pattern indicates Pt(111) as well as Pt-Pb(110) and Pt-Pb(100), demonstrating again that the Pt-Pb HAc nanoplates are composed of a Pt-skin edge and a Pt-Pb alloy core. The line-scan analysis gives more intuitive structural information about the Pt-Pb nanoplates (Fig. S7); that is, the Pt-Pb nanoplates/C allow for a homogeneous distribution of Pt and Pb in the entire area. However, in the case of Pt-Pb HAc/C, the Pb signal is mainly detected in the inner core. To analyze the elemental compositions of the Pt-Pb nanoplates/C and Pt-Pb HAc/C, we performed ICP-AES (Table S1) experiments. The compositional ratio of Pt/Pb in the Pt-Pb nanoplates/C was 1.04, while that of Pt-Pb HAc/C showed a lower Pb content, revealing that the surface Pb leached out during the acid treatment. Based on these characterization results, we conclude that HAc treatment of the Pt-Pb nanoplate leads to the formation of a PtPb/Pt core-shell-like structure.

Electrochemical measurements were conducted using a potentiostat with a three-electrode system. The as-obtained Pt-Pb alloy-based, catalyst-loaded working electrode was pre-cleaned in an Ar-saturated 0.1 mol L^–1 HClO₄ solution for 100 cycles between 0.08 and 1.10 V (vs. RHE) at a scan rate of 100 mV s^–1, and at room temperature. CVs were then recorded in the same potential range at a scan rate of 50 mV s^–1, as shown in Figs. S8a and c. For comparison, we benchmarked their electrochemical properties against those of the commercial Pt/C catalyst (Fig. S8e). The electrochemically active surface areas (ECSAs) were determined by the charge associated with the hydrogen adsorption/desorption regions in the CVs, and they were 8.42, 10.53, and 53.41 m² g^–1 for the Pt-Pb nanoplates/C, Pt-Pb HAc/C, and commercial Pt/C, respectively [32-34]. The ECSA of the commercial Pt/C (particle size of 2–3 nm) was higher than those of the Pt-Pb nanoplates/C and Pt-Pb HAc/C catalysts because the ECSA is strongly associated with the particle size [33]. The relatively large ECSA of Pt-Pb HAc/C as compared to that of the Pt-Pb nanoplates/C indicated a Pt shell-like structure with heterogeneous Pt-Pb distribution after the acid treatment of the Pt-Pb nanoplates. In addition, the CV of Pt-Pb HAc/C showed a similar peak potential for Pt-OH adsorption as that of the commercial Pt/C, indicating the Pt shell-like Pt-Pb HAc/C. Figs. S7b, d, and f show the CO stripping curves of the Pt-Pb nanoplates/C, Pt-Pb HAc/C, and commercial Pt/C, respectively, in Ar-saturated 0.1 mol L^–1 HClO₄ solutions, between 0.08 and 1.10 V, with a scan rate of 50 mV s^–1 at room temperature. The ECSAs were also confirmed by CO stripping, and the obtained values were 8.95, 12.20, and 59.68 m² g^–1 for the Pt-Pb nanoplates/C, Pt-Pb HAc/C, and commercial Pt/C, respectively [32-34]. The onset potential for the electro-oxidation of the adsorbed CO (CO_ads) on Pt was about 0.70 V for the Pt-Pb nanoplates/C catalysts, exhibiting a negative shift compared to that of the Pt-Pb HAc/C (0.78 V) and Pt/C (0.80 V) catalysts. This result demonstrated that Pb on the surface of the Pt-Pb nanoplates changes the electronic structure of Pt and reduces the strength of CO_ads [16, 35]. In addition, the relatively high oxophilic Pb surface sites can provide oxygen-containing species that accelerate the oxidation of CO_ads [35, 36].

The electrocatalytic MOR activity of all the presented catalysts was recorded based on the CV curves from 0.08 to 1.27 V, at a scan rate of 50 mV s^-1, in an Ar-saturated solution containing 0.1 mol L^–1 HClO₄ and 1.0 mol L^–1 CH₃OH. To evaluate the MOR activity, the highest current recorded at a certain potential was normalized with the ECSA and Pt loading, representing specific activity and mass activity, as shown in Figs. 3a and b, respectively. The specific MOR activity of the Pt-Pb nanoplates/C in the forward scan was 12.5 mA cm^–2, which was 1.3 and 18.9 times more than those of the Pt-Pb HAc/C (9.58 mA cm^–2) and commercial Pt/C (0.66 mA cm^–2). The enhanced specific MOR activity of the Pt-Pb nanoplates/C could be ascribed to the homogeneous distribution of the Pt and Pb atoms on the catalytically active surfaces; hence, oxophilic Pb quickly helped Pt site reaction of CO oxidation to CO₂ [16]. The mass MOR activity of the Pt-Pb nanoplates/C at the highest current density in the forward scan was 1.05 A mg^–1, which was comparable to that of the Pt-Pb HAc/C (1.00 A mg^–1) and 2.8 times more than that of the commercial Pt/C (0.37 A mg^–1). Although the specific activity of the Pt-Pb nanoplates/C is higher than that of Pt-Pb HAc/C, the relatively large ECSA of Pt-Pb HAc/C compensates for the activity gap in the Pt-Pb nanoplates/C.

To evaluate the use of Pt-Pb nanocatalysts for practical applications, we conducted long-term durability tests on the three different catalysts in Ar-saturated 0.1 mol L^–1 HClO₄ and 1.0 mol L^–1 CH₃OH solutions, by applying cyclic potential sweeps between 0.6 and 1.1 V, at a scan rate of 100 mV s^–1 for 1000 cycles. Then, the CVs were recorded in a fresh Ar-saturated 0.1 mol L^–1 HClO₄ and a 1.0 mol L^–1 CH₃OH solution between 0.08 and 1.27 V, at a scan rate of 50 mV s^–1 (Fig. 4). After 1000 cycles, the mass MOR activity of the Pt-Pb nanoplates/C, Pt-Pb HAc/C, and Pt/C catalysts decreased to 20.9%, 45.0%, and 21.6% of their initial values, respectively (Fig. 4d). This result was consistent with the current-time curves in Fig. 5. Throughout the test for 3000 s, the residual current density on Pt-Pb nanoplates/C was higher than that of the Pt-Pb HAc/C. As is well known, CO_ads generated as a reaction intermediate blocked the Pt active sites during the methanol electro-oxidation [37-39]. In the case of Pt-Pb nanoplates/C, the modified Pt electronic structure decreased the adsorption energy of CO_ads on the Pt atoms and surface Pb sites, facilitated the adsorption of oxygen-like species, and promoted the oxidation of the adsorbed CO_ads for long reaction cycles [16, 40]. However, Pt-Pb HAc/C with a Pt-Pb/Pt core-shell-like structure provided a smaller amount of oxidants for CO_ads removal; therefore, the CO (poisoning) sites accumulated greatly.

We successfully synthesized Pt-Pb hexagonal nanoplates with a homogeneous distribution on the Pt-Pb alloy surface as highly active and stable electrocatalysts toward the MOR. The specific MOR activity of the Pt-Pb hexagonal nanoplates was about 1.3 and 18.9 times higher than that of the HAc-treated Pt-Pb nanoplates, with a heterogeneous distribution on the Pt-Pb alloy surface and the commercial Pt/C, respectively. After 1000 cycles, the mass MOR activity of the Pt-Pb hexagonal nanoplates showed a 20.9% degradation, revealing more sustainable MOR performance compared to that of the HAc-treated Pt-Pb nanoplates, which showed a 45.0% degradation. After 3000 s of the chronoamperometry MOR test, the residual current density on the Pt-Pb hexagonal nanoplates was 1.4 times more than that on the HAc-treated Pt-Pb nanoplates. The results demonstrated that the Pb distribution on the surface of the Pt-Pb hexagonal nanoplates was a key factor in promoting the MOR activity and stability. The present work is expected to provide a rational approach for the preparation of novel metal nanocrystals that find practical use in DMFCs.

The authors thank Korea Basic Science Institute (KBSI) for the usage of their TEM instrument.