Quanqing
Zhao
a,
Han
Zhi
b,
Liu
Yang
a and
Feng
Xu
*ab
aCollege of Materials Science and Engineering, Fuzhou University, Fuzhou, 350108, China. E-mail: xufeng@fzu.edu.cn
bSchool of Advanced Manufacturing, Fuzhou University, Jinjiang, 362200, China
First published on 28th August 2023
Pt shows excellent catalytic activity and acid stability towards the methanol oxidation reaction (MOR) and oxygen reduction reaction (ORR), but suffers from deactivation due to the weak interaction with the carbon support and the oxidation of carbon under operation. To solve the problem, herein, silica nanospheres were used to load PtCu nanoparticles as the carbon-free catalyst for the MOR and ORR. It is found that the heat treatment is crucial to regulate the metal–support interaction and enhance the catalytic performance. The PtCu/SiO2 catalyst after heat treatment exhibits a specific activity (SA) nearly 7 times the value of Pt/C toward the MOR, and much better durability. The MOR mass activity of heat-treated PtCu/SiO2 is over twice the value of the untreated one. The electron transfer promoted by heat treatment leads to an upshift of the d-band center of Pt, resulting in the increase of the absorption rate of methanol and the intermediates.
Sustainability spotlightDirect methanol fuel cells (DMFCs) are widely accepted as one of the promising green energy resources to replace fossil fuels and are now on their way to commercialization. Pt/C is the commonly used catalyst in DMFCs. The carbon support leads to deactivation during long-term operation. Moreover, carbon is a product of the petrochemical industry. In this work, SiO2 was studied as the non-carbon support to load PtCu alloy. The results showed better performance towards the methanol oxidation reaction and comparable activity towards the oxygen reduction reaction of the PtCu/SiO2. Hence, the activity and durability of the Pt-based catalyst were enhanced and the carbon consumption of DMFCs was further cut, promoting the sustainability of our world. |
Non-carbon support materials, such as titania, alumina, ceria, silica, and tungsten oxide, have received wide attentions recently.5 Most of the oxide materials are inert in an acidic environment and exhibit fascinating advantages over carbon. For example, the surficial groups (such as –OH) on silica promote the MOR through reacting with intermediate species (such as CO*), and reduce the impact of Pt poisoning.6 However, most non-carbon materials exhibit low conductivity which decreases the catalytic activity compared to the carbon support, due to the inherent wide bandgap. In the last decade, it has been revealed that tuning the metal–support interaction (MSI) is effective in tailoring the electronic structure and hence the catalytic performance. MSI engineering includes downsizing the metals, tuning the defects/interface and morphology, and so on.7–9 Fundamentally, the electron transfer involved in the MSI is crucial to the catalytic performance. During electron transfer, excess electrons tend to accumulate at the interface of the support and metal,10 which significantly enhances the adsorption and reactions of adsorbates,11 leading to improved activity and higher conductivity.12 The electron transfer from the support to Pt results in a downshift of the Pt d-band center and consequently weakens the adsorption of methanol and intermediate species,13,14 which promotes the catalytic activity and durability.15
Another urgent issue is the rarity and high cost of Pt, making the DMFCs expensive and hindering their commercialization.16 Using non-noble metals (such as Co, Ni, Cu) to form Pt based alloys is an effective way to solve the issue, and improve electrocatalytic activity.17 Lu18 and co-workers found that Cu could change the binding energy of Pt due to the strong electronic interaction between Pt and Cu. Cu donates electrons and the electron density increases around Pt sites, resulting in weaker chemisorption of intermediates.
Herein, we present a PtCu alloy catalyst supported on silica nanospheres. SiO2, an oxide widely used in industries, was herein studied as a non-carbon support to promote the MSI and hence the activities toward the methanol oxidation and oxygen reduction reactions. Through thermal treatments, the MSI was engineered and the d-band center of Pt was tuned, leading to enhanced MOR performance.
D = Kλ/(βcosθ) | (1) |
Fig. 1 (A) XRD patterns of Cu/SiO2 and HC300; (B) XRD patterns of PtCu/SiO2-40% and HC300-40%; (C) FTIR spectra of SiO2 spheres, PtCu/SiO2-40% and HC300-40%. |
The Fourier transform infrared spectroscopy (FTIR) results of SiO2 nanospheres, PtCu/SiO2-40%, and HC300-40% are exhibited in Fig. 1C. The peaks centered at 800 and 471 cm−1 represent the symmetrical stretching vibration and bending vibration of Si–O. The peaks centered at 949 cm−1 represent the Si–O stretching vibrations of Si–O–H, revealing the rich silanol groups (Si–OH) on the SiO2 surface. The peaks at 1100 cm−1 represent the anti-symmetric stretching vibration of Si–O–Si.23 The peaks at 1633 cm−1 represent the bending or deformation mode of molecularly coordinated water and the peaks at 3424 cm−1 represent the stretching vibration of O–H that were part of the water molecule and Si–OH.24 These peaks were also observed on the other as-prepared catalysts (Fig. S2†). The Pt, Cu, and SiO2 contents were calculated based on the ICP analysis (Table S1†). The Pt, Cu, and SiO2 contents of PtCu/SiO2-40% were 58, 13, and 29 wt% respectively. After heat treatment, HC300-40% showed higher SiO2 content (43 wt%) and lower Pt and Cu contents (46.6 and 10.4 wt% respectively) than PtCu/SiO2-40%, but the Pt:Cu atomic ratio remained almost unchanged (about 1.46). PtCu/SiO2-30% showed higher Pt content but lower SiO2 and Cu contents after heat treatment, and the Pt:Cu atomic ratio became larger (increasing from 0.81 to 1.36). The differences in the content changes on PtCu/SiO2-30% and PtCu/SiO2-40% before and after heat treatment demonstrated that the PtCu alloy of PtCu/SiO2-40% was more stable than that of PtCu/SiO2-30%.
The microstructures of the SiO2 nanospheres, PtCu/SiO2-40% and HC300-40% were investigated using a scanning electron microscope (SEM) and transmission electron microscope (TEM). The solid SiO2 nanospheres were mainly around 100 nm in diameter (Fig. 2A and B). The TEM and EDS images of PtCu/SiO2-40% showed the metal nanoparticles intimately attached on SiO2 nanospheres, thus the PtCu alloy interacted strongly with SiO2 nanospheres (Fig. 2C–G). The metal nanoparticles were mainly 5 nm in diameter (Fig. 2H) and a metal lattice spacing of 0.22 nm was observed, relating to the PtCu (111) facet (Fig. 2I). After heat treatment, most of the PtCu alloy was still loaded on the SiO2 surface and the morphology slightly changed probably because of the heat treatment (Fig. 2J). The measurements of the PtCu (111) crystal plane showed that the spacing of the HC300-40% crystal plane turned into 0.225 and 0.247 nm respectively and the average spacing showed a value of 0.23 nm (Fig. 2K and L), showing that after heat treatment the lattice spacing increased. The elemental mapping (Fig. S3†) of HC300-40% demonstrated the homogeneous dispersion of Pt and Cu elements, and the intimate contact of PtCu and SiO2.
X-ray photoelectron spectroscopy (XPS) was carried out to investigate the electronic structures of Si, Cu, and Pt. The Si 2p orbital of PtCu/SiO2-40% exhibited a peak at around 103.68 eV, and that of HC300-40% exhibited a peak at 103.48 eV (Fig. 3A). There was a negative B.E. shift of the Si 2p after heat treatment, suggesting an increased electron density of HC300-40%. The Cu 2p orbital of PtCu/SiO2-40% showed peaks at 931.88 and 951.68 eV, and that of HC300-40% showed peaks at 932.48 and 952.18 eV (Fig. 3B). The Cu 2p orbital of HC300-40% showed a noticeably positive B.E. shift, suggesting that heat treatment leads to a decreased electron density. The Pt 4f orbital of PtCu/SiO2-40% showed peaks at 71.38 and 74.58 eV, and that of HC300-40% showed peaks at 71.68 and 74.88 eV (Fig. 3C). There was also a positive shift in Pt 4f binding energy after heat treatment, suggesting a decreased electron density. The HC300-30% showed similar shifts of Pt 4f, Cu 2p and Si 2p orbitals after heat treatment compared to HC300-40% (Table S2†). The positive shifts of Pt 4f and Cu 2p and the negative shift of Si 2p revealed that even though Cu donated electrons to Pt, the electron density of Pt still decreased and the electrons transferred from PtCu to SiO2, which demonstrated a strong MSI.7
Fig. 3 XPS spectra of the Si 2p (A), Cu 2p (B), and Pt 4f (C) orbitals of PtCu/SiO2-30%, HC300-30%, PtCu/SiO2-40% and HC300-40%. |
The MOR activity of PtCu/SiO2-10%, PtCu/SiO2-20%, PtCu/SiO2-30%, and PtCu/SiO2-40% is compared in Fig. S3.† As the Pt loading increases, the activity rises and the PtCu/SiO2-40% shows much higher activity than the others; therefore, the discussion hereafter focuses on PtCu/SiO2-40% and HC300-40%. The ECSA was calculated from CV curves (Fig. 4A) using a hydrogen adsorption theory by assuming a monolayer adsorption of hydrogen atoms on the surface of metallic catalysts (0.21 mC cm−2).25 The ECSA of Pt/C (30.8 m2 gPt−1) was higher than that of PtCu/SiO2-40% (3.53 m2 gPt−1) and HC300-40% (4.4 m2 gPt−1) because the Pt nanoparticles of Pt/C exhibited smaller size (<5 nm) and better dispersion, and the electronic conductivity of the carbon support is higher.26 The Nyquist plot illustrated that although the electronic resistance of SiO2 nanospheres was higher than that of carbon, it was low enough as a support (Fig. S4†). The PtCu/SiO2-40% demonstrated even lower resistance than SiO2, and the resistance was further reduced after heat treatment, which evidenced the strong MSI.
Relative to Pt/C, PtCu/SiO2-40% showed poor MOR activity, but HC300-40% exhibited comparable activity (Fig. 4B). The mass activities (MAs) of HC300-40% (0.28 A mgPt−1) and Pt/C (0.29 A mgPt−1) were very close (Fig. 4C); the MA of PtCu/SiO2-40% was 0.13 mgPt−1, less than half that of HC300-40%. The HC300-30% also showed an increase of MOR activity compared to PtCu/SiO2-30% (Table S3†).
The ratio of peak current densities of the forward scan (If) to backward scan (Ib) is commonly used to identify the poison resistance of the catalysts.27 In the forward scan, methanol was oxidized into intermediates which strongly adsorb on the active sites, then in the backward scan the intermediates resume being oxidized.28 Higher If:Ib ratios indicate that more intermediates are oxidized in the forward scan, leaving less to the backward scan. Compared to the If:Ib ratio of 0.93 (Pt/C), PtCu/SiO2-40% and HC300-40% demonstrate higher values of 1.30 and 1.27. Obviously, the poisoning resistance of the PtCu/SiO2 catalysts is improved significantly. The specific activities (SAs) of PtCu/SiO2-40% and HC300-40% are 3.66 and 6.35 mA cm−2 respectively, and the SA of HC300-40% is nearly 7 times the value of Pt/C (0.95 mA cm−2).
The durability of the catalyst is a highly important issue and was also tested (Fig. 4D). The ECSA of Pt/C decreased significantly in the first 1000 cycles, which meant poor durability due to the weak metal–support interaction and the oxidation of the carbon support.29 The PtCu/SiO2-40% showed the best durability with 39.49% ECSA remaining after 5000 cycles, and an ECSA of 8.14% for HC300-40% remained after 5000 cycles. PtCu/SiO2-40% and HC300-40% retained 32.7% and 38% of their MOR activity and all catalysts retained more than 30% after 5000 cycles, and the If:Ib ratios of PtCu/SiO2-40% and HC300-40% were 1.26 and 1.24 respectively (Fig. S5†). Such slight decreases after ADT demonstrated the minor decline of poisoning resistance.
The analysis of electrochemical performance showed that the heat treatment was crucial to regulate the MSI between the SiO2 and PtCu alloy and improve the MOR activity. The If:Ib ratios of PtCu/SiO2-40% and HC300-40% were 1.26 and 1.24 respectively (Fig. S4†). Such slight decreases after ADT demonstrated the minor decline of poisoning resistance. The electron transfer promoted by heat treatment led to an upshift of the d-band center of Pt,7 which meant easier charge donation from Pt to the adsorbates and the chemisorption on Pt became stronger, leading to the increase of the absorption rate of methanol and the intermediates.30–32 In the durability test, the ECSA of HC300-40% drops more than that of PtCu/SiO2-40%, owing to the increase of the absorption rate of the intermediates such as CO*. Meanwhile, the HC300-40% showed less MOR activity decrease than PtCu/SiO2-40% after ADT, due to the MSI which promoted charge accumulation at the interface during electron transfer and increased the specific activity of Pt active sites, and the existing –OH at the interface protecting the Pt active sites from poisoning. Thus, the electron transfer caused by heat treatment led to much better MOR activity and the durability also improved.
Many studies have shown that the shift of the d-band center of Pt can change the ORR activity.21,33,34 The ORR results (Fig. 5A) illustrated that PtCu/SiO2-40% and HC300-40% exhibited close onset potentials to Pt/C, but much better durability. After 10000 cycles, the ORR activity of PtCu/SiO2-40% became better and that of HC300-40% remained unchanged, while that of Pt/C showed a significant decline (Fig. 5B–D). The PtCu nanoparticles strongly bonded to SiO2 nanospheres due to the MSI, leading to improved ORR durability. The structural evolution of PtCu during the ORR was the reason for the better activity after ADT. Besides, the ascorbates were wiped out to form a clean surface during operation, leading to better activity.35
50 mg of Cu/SiO2 was placed in a tube furnace and calcined at 300 °C for 2 hours in an atmosphere of 10 vol% H2/N2. The obtained sample was named HC300, loaded with 30 wt% and 40 wt% Pt and treated with nitric acid solution. Then the solution was filtered and the obtained samples were washed three times with water and lyophilized, and named HC300-30% and HC300-40% respectively.
The GCE was finely polished to a mirror-like surface with an Al2O3 slurry (<50 nm) before use. Catalyst ink was prepared by mixing 2 mg of catalyst, 0.5 mL of isopropanol, 0.5 mL of deionized water and 10 μL of Nafion with ultrasonic dispersion. 10 μL of ink was dropped on the working electrode and dried in the air. Cyclic voltammetry (CV) was performed in a nitrogen saturated 0.5 M H2SO4 electrolyte in the potential range of 0–1.2 V vs. the reversible hydrogen electrode (RHE) at the scanning rate of 100 mV s−1. The MOR activity test was performed in a nitrogen saturated solution containing 0.5 M H2SO4 and 1 M methanol, and the CV program was selected for the MOR activity test. The MOR ADT was performed in 0.5 M H2SO4 electrolyte in the potential range of 0.6–1.2 V vs. RHE at the scanning rate of 50 mV s−1. Before the ADT and after every 1000 cycles, the MOR activity test was performed and recorded.
For the ORR tests, CV was performed in O2 saturated 0.1 M HClO4 aqueous solution in the potential range of 0–1.2 V vs. RHE at the scanning rate of 50 mV s−1. Linear scanning voltammetry (LSV) was performed in O2 saturated 0.1 M HClO4 aqueous solution in the potential range of 0–1.2 V at the scanning rate of 10 mV s−1 with the rotating speed of 1600 rpm. The ADT was completed by performing CV in N2 saturated 0.1 M HClO4 aqueous solution in the potential range of 0.6–1.2 V at the scanning rate of 100 mV s−1 for 5000 cycles.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d3su00193h |
This journal is © The Royal Society of Chemistry 2023 |