Pt nanoparticles supported on nitrogen-doped porous carbon as efficient oxygen reduction catalysts synthesized via a simple alcohol reduction method

Pt nanoparticles supported on nitrogen-doped porous carbon (NPC) were investigated as both a highly active catalyst for the oxygen reduction reaction (ORR) and a suitable porous support structure. Pt/NPC catalysts with loadings of 8.8–35.4 wt.% were prepared via a simple alcohol reduction method and exhibited homogeneously dispersed Pt nanoparticles with a small mean size ranging from 1.90 to 2.99 nm. X-ray photoelectron spectroscopy measurement suggested the presence of strong interactions between the Pt nanoparticles and NPC support. 27.4% Pt/NPC demonstrated high catalytic activity for the ORR in a rotating disk electrode system and was also effectively applied to a gas diffusion electrode (GDE). A GDE fabricated using the Pt/NPC with a fine pore network exhibited excellent performance, especially at high current densities. Specific activity of Pt/NPC and Pt/carbon black catalysts for the ORR correlated with the peak potential of adsorbed OH reduction on Pt, which was dependent on the particle size and support.


Introduction
Electrocatalysts for the oxygen reduction reaction (ORR) in alkaline media have attracted tremendous interest for their applications in alkaline fuel cells [1][2][3][4], metal-air batteries [5,6], and chlor-alkali electrolysis with oxygen depolarized cathodes [7]. Recently, alkaline membrane fuel cells, which can be simplified and the CO 2 poisoning problem for alkaline fuel cells is alleviated [8], have been reported to exhibit improved performance, which approaches that of proton exchange membrane fuel cells, due to advances in anion-exchange membranes [9]. At present, Pt and Ptalloy nanoparticles supported on carbon are universally available electrocatalysts for the ORR [10,11]. Among carbon support materials, carbon blacks (CBs) are commonly used in gas diffusion electrodes (GDEs) for fuel cells and metal-air batteries because of their large surface area, good electric conductivity, and well-developed porous structure as well as low cost [12,13]. Despite these advantages, the slow kinetics of CB-supported Pt nanoparticles for the ORR necessitates high loading levels of Pt, hindering the further deployment of the electrochemical energy conversion and storage devices.
In order to improve the performance of Pt nanoparticle catalysts, decreasing the particle size and thereby increasing the surface area per unit mass is one of the reasonable strategies in heterogeneous catalyst fields [14]. However, many reports have demonstrated that the specific activity (current density per unit surface area) of Pt nanoparticles for the ORR decreases with decreasing particle size between about 1 and 5 nm and mass activity (current density per unit mass) does not linearly increase in this size range [15][16][17]. The effects of small Pt size on the ORR activity may be predicted by considering the ratio of terrace, corner, and edge sites on the particles [18,19]. Following the geometrical considerations, the distribution of the surface sites rapidly changes for the small nanoparticles, which may lead to the strong adsorption of oxygen species, such as OH and O, through an increased number of low-coordinated sites [20,21]. The strongly adsorbed species on the surface inhibit the adsorption of molecular oxygen and intermediates for the ORR [22]. Thus, the state and coverage of the oxygen species is considered to be a key factor in determining the activity of Pt catalysts [13,23,24]. Besides, it is suggested that the ORR activity of Pt nanoparticle catalysts can be related to interparticle distances and mass transport of oxygen [25,26]. Watanabe et al. have reported that, if the Pt interparticle distance is too close (< 20 nm), the diffusion field of the particles for O 2 overlaps, resulting in decreased specific activity for the ORR [25].
It is widely recognized that support materials can significantly affect the catalytic activity of Pt nanoparticles [27][28][29]. Therefore, long-running efforts to develop various supports such as graphene [30,31], carbon nanotubes [32,33], CBs [34,35], metal carbides [36,37], and metal oxides [38] have been undertaken. Carbon supports with high graphitic nature have both excellent electrical conductivity and high mechanical strength, but the carbon lattice with strong sp 2 chemical bonds does not readily permit the chemisorption of Pt on the basal planes [39]. Because nitrogen doping can introduce anchoring sites for Pt deposition, nitrogen-doped carbon has been used as a support for Pt catalysts over the past few decades [40][41][42][43]. Furthermore, it is suggested that surface nitrogen species such as pyridinic N in nitrogen-doped carbon increase the interactions between Pt catalysts and nitrogen-doped carbon supports, which may enhance the catalytic performance [42,44,45]. Nitrogen-doped porous carbon (NPC) supports with large surface area and high pore volume, which can act as a substrate for the effective dispersion of Pt nanoparticles and facilitate mass transport, would contribute to improving the ORR activity of Pt nanoparticles.
In the present study, Pt nanoparticles supported on NPC were prepared via a surfactant-free alcohol reduction method and characterized by transmission electron microscopy (TEM), X-ray photoelectron spectroscopy (XPS), X-ray diffraction (XRD), scanning electron microscopy (SEM), and N 2 gas adsorption. To evaluate the ORR activity, rotating disk electrode (RDE) measurements were conducted on the synthesized catalysts. Furthermore, a GDE was fabricated using Pt/NPC as an ORR catalyst and exhibited excellent performance, especially at high current densities.

Preparation of Pt nanoparticles supported on nitrogen-doped porous carbon
NPC was prepared via heat treatment using commercial CB (Ketjen Black EC600JD, Lion) and cyanamide (Sigma-Aldrich), which were used as carbon and nitrogen sources, respectively (for full details, see supplementary material) [46]. Pt nanoparticles supported on NPC were synthesized via an alcohol reduction process [47,48] using dinitrodiamine platinum as a platinum source and methanol as a solvent and reducing agent. In a typical procedure, 100 mg of NPC with a nitric acid solution of dinitrodiamine platinum (Kojima Chemicals) was suspended by sonication in 10 mL of methanol for 10 min. The Pt loading was controlled by varying the amount of dinitro-diamine platinum. The mixtures were then heated to reflux under N 2 flow for 6 h in a round-bottom flask equipped with a condenser, thermal controller, and magnetic stirring bar. The reduction process was monitored by observing the color changes in the solution; from the original transparent yellow to colorless (Fig. S1a, b). In addition, the UV-Vis absorption spectra of the solutions were measured using a UV-Vis-NIR spectrophotometer (UV-3600, Shimadzu). The absorbance in wavelengths lower than 470 nm with a shoulder at 260 nm of dinitrodiamine platinum solution was not observed in the spectra for the solution after 6 h of methanol reduction (Fig. S1c). After the mixtures cooled, the solid precipitates were filtered, washed thoroughly with ethanol and distilled water, and then fully dried at 40 °C for 12 h. Pt nanoparticles supported on CB as reference catalysts were prepared in the same manner.

Material characterization
The Pt content in the synthesized catalysts was confirmed by inductively coupled plasma atomic emission spectroscopy (SPECTRO ARCOS, SPECTRO). The morphology of samples was observed by TEM (Titan Cubed G2 60-300, FEI Company) and SEM (SU-8220, Hitachi High-Technologies). The sample powders for TEM measurements were suspended in ethanol and dropped onto holey carbon films supported by Cu microgrids. The interparticle distance was calculated by where X is the interparticle distance in nm, σ is density of catalyst particles in g nm −3 , d is mean size of catalyst particles in nm, S is specific surface area of supports in nm 2 g −1 , and y is catalyst content in wt.% [25,49]. XPS (Quantera SXM, Ulvac-phi) was performed using an Al Kα X-ray source (1486.6 eV). The peak shift correction was based on the binding energies of Cu 2p at 932.62 eV and Au 4f at 83.96 eV. Sample powders for XPS measurements were uniformly pressed onto an indium plate (Nilaco). Fitting of the N 1 s peak was performed using four common bonding configurations within the carbon lattice, including pyridinic N, pyrrolic N, quaternary N, and oxidized N [50]. The Pt 4f peak was fitted using the 4f 7/2 and 4f 5/2 components, which were separated by 3.3 eV with a fixed area ratio of 4/3 [51]. XRD patterns were obtained using an X-ray diffractometer (X'Pert Pro, Spectris) with a Cu Kα radiation source operated at 45 kV and 35 mA. The crystallite size of Pt in nm was calculated using the Scherrer's equation: where L is the crystallite size in nm, K is the dimensionless shape factor (0.9), λ is the wavelength of Cu Kα radiation (0.154 nm), β is the full-width half-maximum of Pt (220) peaks in radians, and θ is the angle at the position of the peak maximum in radians [52]. Raman data were collected using a confocal Raman microscope (InVia Reflex, Renishaw) with a 532 nm laser as the excitation source. N 2 adsorption/desorption isotherms were obtained using a surface area and pore size distribution analyzer (BELSORP-max, Nihon BEL). The pore size distribution was estimated on the basis of the Barrett-Joyner-Halenda [53]. Specific surface area was calculated from the Brunauer-Emmett-Teller (BET) equation in the partial pressure range of 0.05 − 0.20 N 2 [54]. The micropore volume was determined from the intercept of the t-plot [55]. The total pore volumes were obtained from the total adsorbed gas amount at a partial pressure of 0.98. The conversion factor from the amount adsorbed to the volume of liquid adsorbate was 0.0015468 under the assumption that the density of the condensed adsorbate in the pores is equal to the density of the bulk liquid adsorbate [56].

Electrochemical characterization
Electrochemical measurements were performed using a RDE and a GDE. All water used was first purified by a Barnstead Nanopure Diamond System (18.2 MΩ cm resistivity). Potentials were iR-corrected to compensate for the effect of solution resistance. Working RDEs (0.196 cm 2 geometrical surface area) were prepared as follows: initially, 5.0 mg of a catalyst sample was dispersed in 2-propanol solvent (Wako Chemical) with 50 µL of a 5.0 wt.% Nafion solution (Sigma-Aldrich). The mixture (5.0 mg mL −1 ) was agitated ultrasonically for 30 min in an ice bath to form a uniform catalyst ink. The suspension (3.3 µL) was deposited in increments of 1.1 µL onto the electrode on an inverted rotator shaft (gently spinning at 0-100 rpm) and dried while spinning at ~ 100 rpm in air at room temperature for ~ 5 min, producing Pt loadings on the RDEs of 7.4-29.8 µg Pt cm −2 for the catalysts with Pt concentrations of 8.8-35.4 wt.%. The catalyst-deposited electrode was further dried at 60 °C for 30 min to fully remove the 2-propanol. Electrochemical measurements were conducted using a potentiostat (HZ-5000, Hokuto Denko) with a three-electrode cell system. The reference electrode was an Hg/HgO electrode (KOH, 0.1 M), the counter electrode was a platinum wire, and the electrolyte was 0.1 M KOH solution. Before electrochemical measurements, ~ 100 cycles of cyclic voltammetry were performed by scanning the potential between 0.3 and − 0.8 V vs. Hg/HgO at a sweep rate of 0.5 V s −1 in a N 2 -saturated electrolyte to clean the surfaces of the catalysts. The background currents were measured at a sweep rate of 20 mV s −1 in the N 2 -saturated electrolyte. After the background current was measured, the electrolyte was saturated for more than 30 min with O 2 . Linear sweep voltammograms were recorded at a sweep rate of 20 mV s −1 at 25 °C, and the oxygen reduction current was corrected by the background current. For the Tafel plot, the kinetic current density (j k ) was calculated from a mass-transport correction of the RDE by where j is the disk current density after correction for background current and j l is the diffusion-limiting current density [57,58]. Mass and specific activities were estimated via calculation of the kinetic current and normalization to the Pt loading and electrochemical surface area, respectively. The electrochemical surface area was calculated using Eq. (4) on the basis of the hydrogen adsorption charge in the negative potential scan from 0.40 to 0.05 V vs. RHE at a sweep rate of 20 mV s −1 on cyclic voltammograms after the current at 0.40 V was subtracted from the total current to correct for double-layer charging.
where Q H is the measured hydrogen charge, W Pt is the loading of Pt, and 0.21 mC cm −2 is the charge required to oxidize a monolayer of hydrogen on the Pt sites [57,59]. Because the current at ~ 0.4 V includes a significant contribution from the surface (100) terrace domains on the Pt nanoparticles in alkaline solution, the electrochemical surface area of the synthesized catalysts was determined in 0.1 M HClO 4 in order to avoid overestimating the double-layer charging [60,61].
Polarization curves of the GDEs with incorporated catalysts were obtained galvanostatically. The GDEs were fabricated using a previously reported procedure [62]. The GDE (diameter = 14 mm) comprised a catalyst layer (about 0.1 mm thick) and a gas diffusion layer (about 0.1 mm thick) formed on a Ni mesh current collector (100 mesh, Nilaco). The sample powder, a polytetrafluoroethylene (PTFE) dispersion (D-210C, Daikin Industries), and 1-butanol (Kishida Chemical) were added to distilled water and agitated ultrasonically for 30 min. The PTFE content in the catalysts was fixed at 30 wt.%. After filtration and drying at 125 °C, a powder was obtained for the catalyst layer. The powder for the gas diffusion layer was fabricated separately by mixing acetylene black (mean diameter = 48 nm; HS-100, Denki Kagaku Kogyo), a PTFE dispersion, and Triton-X surfactant (Kishida Chemical) in distilled water. The PTFE content of this powder was also fixed at 30 wt.%. After the mixture was filtered and dried at 125 °C, the obtained powder was heated at 280 °C for 3 h in an electric furnace to remove the surfactant. The powders for both layers and the Ni mesh were hot-pressed at 360 °C to produce the GDEs. The steady-state current density-potential curves of the PTFE cells installed with the fabricated GDEs were measured in 8

Material characterization of Pt nanoparticles supported on nitrogen-doped porous carbon
To determine the mean size of the Pt nanoparticles in synthesized catalysts, they were characterized by TEM (Fig. 1) (Fig. 1f) shows clear lattice fringes throughout the particles, with a distance of 0.23 nm, which is consistent with the d-spacing of (111) planes of face-centered-cubic Pt, corroborating that the nanoparticles are Pt nanocrystals. Figure 2A, B and Table 2 show the C 1 s and N 1 s XPS spectra and atomic concentrations of NPC and CB supports. The XPS overview spectra of Pt/NPCs, Pt/CB, NPC, and CB are shown in Fig. S2. The C 1 s peak of sp 2 carbon at 284.5 eV for NPC is asymmetric and broad toward the high-binding-energy side due to the presence of C-N, C-O, C=O, O=C-O, and plasmon bands [63,64]. Nitrogen atoms serve to mediate the enhancement in the strong adsorption of Pt by activating nitrogen-neighboring carbon atoms [65,66]. Such activation can facilitate nucleation and subsequent particle growth. The increased number of nucleation sites      [67,68]. First-principles study revealed that Pt nucleation is favorable on pyridinic and pyrrolic N, and the center of d-band can be shifted downward by the nitrogen species [69]. Thus, the species on the NPC could be attributed to the increase in interactions between the Pt nanoparticles and NPC support. The interactions would enhance adhesion of the Pt nanoparticles to the NPC support [70], which could alleviate the aggregation of the nanoparticles. The XRD spectra of the Pt/NPCs, 30.3% Pt/CB, and NPC are shown in Fig. 3. Two broad peaks for NPC were observed at about 24.5° and 43.5°, corresponding to the (002) and (101) diffraction peaks of hexagonal graphitic carbon (JCPDS No. 75-1621). The broad (002) basal plane peak indicates that the NPC comprises small crystalline structures. The diffraction peaks of the face-centeredcubic Pt can be identified from the spectral database (JCPDS No. 04-0802). The crystallite sizes of the Pt in 19.6% Pt/NPC, 27.4% Pt/NPC, 35.4% Pt/NPC, and 30.3% Pt/CB were calculated to be 2.11, 2.48, 2.82, and 3.14 nm, respectively, from the (220) peaks using Scherrer's equation. The increase in the crystallite size of the Pt/NPCs with increasing Pt loading is consistent with the increasing trend of the mean particle sizes revealed by TEM. The mean particle size of 2.13 nm for 19.6% Pt/NPC was almost the same as the crystalline size of 2.11 nm, indicating that the Pt nanoparticles were monocrystalline. However, the mean particle size of 2.57 and 2.99 nm for 27.4% and 35.4% Pt/NPCs were slightly larger than the crystalline sizes of 2.48 and 2.82 nm, respectively, suggesting that some Pt nanoparticles were aggregated on the NPC support. The interparticle distance of the Pt/NPCs increased from 22.6 to 25.1 nm as the loading increased from 19.6% to 35.4%, suggesting a decrease in the number of particles on the NPC, which can be attributed to the slight aggregation of the nanoparticles.
Figs. S3, and S4 show SEM images and Raman spectra of 27.4% Pt/NPC, 30.3% Pt/CB, NPC, and CB. The Raman spectra consist of two dominant peaks assigned to the first order G and D modes. The G mode has E 2g symmetry and is attibuted to the inplane stretching vibration of carbon atom pairs [71]. The D mode has A 1g symmetry and is attributed to the inplane breathing vibration of the six-membered aromatic carbon rings [71]. The two additional broad peaks around 1200 and 1515 cm −1 correspond to the disordered graphite similar to polyaromatic hydrocarbons and a graphitic layer containing distorted structures such as integrated heteroatoms [72,73]. The size of NPC primary particles and intensity ratio of D (disorder) and G (graphite) bands, I D /I G , for 27.4% Pt/ NPC and those for NPC were found to be nearly identical: about 30 nm and 1.34, respectively, and the deposition of Pt nanoparticles on NPC did not significantly affect 20  b a the nanostructure. Moreover, the particles appear to be porous, which is advantageous for their application in fuel cells and metal-air batteries [74,75]. The primary particles form agglomerates, and mesopores are present between the particles within the agglomerates. Furthermore, the agglomerates form chain-like aggregates, creating both an extensive network of meso-and macropores and continuous electronic conducting pathways (schematic illustration shown in Fig. S5). The N 2 adsorption/desorption isotherms and BJH pore size distributions for the Pt/NPCs, NPC, 30.3% Pt/CB, and CB are shown in Figs. 4 and S6. The BET surface area, total pore volumes, and micropore volumes are summarized in Table S1. The isotherm of the NPC displays a distinct hysteresis loop over a wide partial pressure region (P/P 0 > 0.42), which is attributed to mesopores. Near a P/P 0 value of 1.0, no saturation was observed, indicating the presence of macropores. Although the adsorption volume per mass was reduced along with increasing weight percentage of Pt, the characteristic shapes of isotherms were also obtained for all Pt/NPCs.

Electrocatalytic properties of Pt nanoparticles supported on nitrogen-doped porous carbon
To evaluate the electrocatalytic activity of the synthesized catalysts for the ORR, RDE measurements were conducted in O 2 -saturated 0.1 M KOH solution. Voltammograms for the Pt/NPCs and 30.3% Pt/CB show a single reduction wave with a mixed kinetic-diffusion control region, about 0.71 V < E < 0.95 V, above a flat diffusion-limited current density plateau at approximately − 5.7 mA cm −2 (Fig. 5a). The current density value is consistent with the four-electron limiting current for the ORR, which was previously reported for a platinum single-crystalline electrode [76] and corresponds to a theoretical value calculated using the Levich equation, suggesting a negligible contribution from O 2 diffusion through the Nafion binder [57,77]. In order to quantitatively determine the ORR pathway, rotating ring disk electrode measurements were further performed (Fig. S7). The number of electrons transferred was approximately four, and almost no hydrogen peroxide formation was observed for the Pt/NPCs and 30.3% Pt/CB in the potential region above 0.57 V. Figure 5b shows Tafel plots of the RDEs for synthesized catalysts. The slopes were − 69.7, − 62.0, − 63.8, − 59.2, and − 62.6 mV decade -1 for 8.8% Pt/NPC, 19.6% Pt/NPC, 27.4% Pt/NPC, 35.4% Pt/NPC, and 30.3% Pt/CB, respectively. The slope values of approximately − 60 mV decade −1 for the Pt/NPCs (excluding 8.8% Pt/NPC) and 30.3% Pt/CB would be ascribed to Temkin adsorption, in which coverage of oxygen species, such as OH and O, on the Pt surface was high [78,79]. OH would be the main oxygen species in alkaline solution [80]. The kinetic current density and mass   [81], but it is higher than 0.028 A mg Pt −1 for 12.9% Pt/nitrogen-doped graphene (mean size: 6.2 nm, nitrogen concentration: ~ 5 wt.%) [82] and 0.075 A mg Pt −1 for commercial 10% Pt/C (Johnson Matthey) [81]. To investigate the usage of synthesized Pt/NPC, 27.4% Pt/NPC was used as an ORR catalyst in a GDE; GDEs have been utilized as oxygen electrodes in fuel cells and metal-air batteries [74,75]. Figure 5d shows steady-state polarization curves for GDEs using 27.4% Pt/NPC and 30.3% Pt/CB. The kinetic activity cannot be directly read from the iR-corrected plots for GDEs, at high current densities where mass transport effects are significant, whereas at lower current densities the cathode reaction kinetics become more dominant in determining the GDE performance [74,83]. No obvious potential drop was observed in the measured current density in the Tafel plot of 27.4% Pt/NPC, suggesting that the plot shape is kinetically controlled (Fig. 5d inset). Moreover, the slope values of the Tafel plots were approximately − 60 mV decade −1 (− 59.3 and − 57.5 mV decade −1 for 27.4% Pt/NPC and 30.3% Pt/ CB), consistent with the RDE results. In GDEs for alkaline fuel cells and metal-air batteries, oxygen molecule is reduced to hydroxide at the three-phase boundary between gas, the alkaline solution, and the solid catalyst [74]. For optimal utilization of the three-phase boundary, meso-and macroporous structures of support materials are particularly important because the large pores can provide effective paths of mass transfer in the catalyst layer [84,85]. Thus, the high meso-and macropore volumes, which constitute 89% of the total pore volume (Table S1), as well as good electronic conducting pathways in the porous structure of 27.4% Pt/NPC would be favorable for use in a GDE. In particular, at high current densities, not only the high catalytic activity for the ORR but also the porous structure of 27.4% Pt/NPC may contribute to its superior electrode performance. For comparison, the potentials of GDEs at current densities of 100 and 250 mA cm −2 are summarized in Table S2. In terms of the potentials, the performance of the GDE using 27.4% Pt/NPC is superior than that of a previously reported GDE using commercial Pt/C (TEC 10A30E, Tanaka Kikinzoku Kogyo) installed in a PTFE-made cell, which was also used in the present study [62], and other GDEs in alkaline solutions.
We further compared the specific activity and summarized the electrochemical results for the synthesized catalysts in Fig. 6a and Table S3. A correlation was observed between the mean particle size and the activity of the Pt/NPCs, although 30.3% Pt/CB with larger mean particle Fig. 6 a Potential with respect to specific activity of Pt/NPCs and 30.3% Pt/CB. For the plots, the Tafel plots on RDE (Fig. 5b) were replotted. The specific activity was calculated by normalizing the kinetic current to the electrochemical surface area (Fig. S8) (Fig. 6b). Figure 6c and d show peaks associated with the reduction of the adsorbed OH (OH ad ) on the Pt nanoparticles and the dependence of the specific activity on the peak potential for the synthesized catalysts, respectively. The potential is a good indicator of the adsorption strength of oxygen species, which inhibit adsorption of molecular O 2 , on Pt [86][87][88]. Lower potentials for Pt/NPCs with smaller mean particle sizes would indicate stronger adsorption of the species. Therefore, the lower specific activity of Pt/ NPCs with smaller mean particle sizes would be attributed to the stronger adsorption of the oxygen species on the nanoparticles. Mayrhofer et al. found a similar negative shift of OH ad reduction peaks with decreasing particle size, as one may expect for flatter surfaces and a decreasing average atomic coordination number [21]. Thus, the decrease in the specific activity of Pt/NPC with smaller mean particle sizes could be attributed to the decrease in the average coordination number. Notably, 35.4% Pt/NPC with small mean particle size of 2.99 nm exhibits a slightly high peak potential of OH ad reduction and improved activity compared with 30.3% Pt/ CB (mean particle size: 3.24 nm). Pt nanoparticles supported on nitrogen-doped carbon are typically prepared via deposition of the nanoparticles from platinum salts [89]. In the synthesis, Pt nanoparticles would nucleate and grow on the nitrogen atoms, where the Pt nanoparticles and nitrogen-doped carbon support can strongly interact [45,65]. A positive shift of the binding energy for Pt 4f in the XPS spectra of the Pt/NPCs would reflect a modified electronic structure of Pt by NPC support, which can decrease the adsorption strength of detrimental oxygen species, such as OH and CO, on Pt [68]. Thus, the decrease in the adsorbate-Pt strength might enhance the specific activity of Pt nanoparticles supported on NPC. Although a detailed mechanism that underlies the enhancement has to be studied further, the present study demonstrates an efficient structure that could guide the design of improved catalysts using Pt nanoparticles and NPC.

Conclusions
Pt nanoparticles supported on highly porous nitrogen-doped carbon were prepared via a simple alcohol reduction method. Pt nanoparticles with mean size in the range from 1.90 to 2.99 nm were uniformly deposited on NPC support with loadings of 8.8-35.4 wt.%. The increased number of nucleation sites provided by nitrogen atoms would contribute to the deposition of small nanoparticles. Pt 4f XPS spectra showed a positive shift of the binding energy, suggesting the presence of strong interactions between the Pt nanoparticles and NPC. Although 8.8% Pt/NPC with too small particle size exhibited low mass activity for the ORR, the Pt/NPCs (excluding 8.8% Pt/NPC) showed superior mass activity compared with 30.3% Pt/CB. When used as an ORR catalyst in the construction of a GDE, 27.4% Pt/NPC with high catalytic activity and hierarchical pore structure exhibited no obvious potential drop in its Tafel plot. The specific activity for the ORR of the Pt/NPCs and 30.3% Pt/ CB correlated with the peak potential of OH ad reduction, reflecting the adsorption strength of oxygen species on Pt, which was dependent on the mean particle size and support.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.