Enhanced Durability and Catalytic Performance of Pt–SnO2/Multi‐Walled Carbon Nanotube with Shifted d‐Band Center for Proton‐Exchange Membrane Fuel Cells

Worldwide, significant efforts are made to identify energy sources that can help achieve carbon neutrality and promote sustainable development. The development of a catalyst that combines durability and high performance is essential for the commercialization of proton‐exchange membrane fuel cells (PEMFCs). In a fuel cell, carbon corrosion occurs during startup and shutdown due to improper local flooding caused by inadequate water management. In this study, a Pt‐based catalyst is designed with excellent durability and high activity. Introducing a metal oxide layer modified with Pt/multi‐walled carbon nanotubes reduces the direct contact between carbon and the fuel cell environment. This helps prevent carbon corrosion and inhibits the separation, aggregation, and growth of Pt nanoparticles. Moreover, the catalyst exhibits enhanced oxygen reduction activity due to the electronic effect of the metal oxide layer that is coated on it. In this study, by implementing a carbon erosion acceleration protocol, excellent catalytic properties during a load‐cycling experiment consisting of 5,000 cycles are reported. The practical application of the developed catalyst in PEMFCs offers an effective approach to developing Pt‐group metal catalysts with exceptional activity.

increase performance by increasing the reaction rate of the catalyst.Their study reported that the binding energy of the d-band center should be lowered because the kinetic reaction rate is slow owing to the high affinity for oxygen because of the high energy of the d-band center in the ORR mechanism.When the d-band center of Pt is high, it is located at an energy closer to the Fermi level.Moreover, owing to the high d-band center, Pt has a stronger binding energy with oxygen, surpassing the required strength.Therefore, Pt slows the reaction rate with oxygen in the ORR.One solution to this problem is to reduce the binding energy between Pt with oxygen. [27]To reduce the binding energy, studies have suggested a change in the alloy form of Pt [28,29] or a decrease in the d-band due to the influence of other metal factors. [30]

Carbon Corrosion
Carbon is thermodynamically unstable under the operating conditions of PEMFC and undergoes corrosion via oxidation. [31]hen the carbon support is corrupted, the interaction between the Pt NPs and carbon support is weakened.The electrochemical surface area (ECSA) of Pt is reduced, thereby reducing the life span and performance of the PEMFC catalysts. [32]In general, the carbon oxidation kinetics are negligible under normal operating conditions.However, certain conditions such as the presence of water, low pH, high temperature, and overvoltage during fuel cell operation cause carbon support corrosion.[35] Generally, carbon support corrosion occurs predominantly at the cathode as follows (Equation ( 1)).
Carbon corrosion initiates at 0.207 V RHE , but the reaction rate is so slow that it could be ignored under the normal operating conditions of the PEMFC.However, corrosion of the carbon support of the fuel cell catalyst occurs during the necessary processes, such as fuel starvation and on-off operation (Figure 1).Several studies have been conducted to address the carbon corrosion problem.Graphene oxide, [36] a carbon support with a crystalline structure, [37,38] has been studied as an alternative to existing carbon supports.In addition, non-carbon supports, such as metal oxides, [39,40] carbides, [41] and nitrides, [42] have been studied.To overcome carbon corrosion, the structure and composition of the catalyst and GDLs are being improved, and new concepts for catalysts with higher catalytic activity and stability are also being developed.These studies contribute to improvements in the durability and performance of PEMFCs.
Many approaches have been proposed to overcome the problems caused by carbon corrosion in catalysts and enhance their catalytic performance.Carbon black has been commercialized and used as an electrode catalyst support in low-temperature PEMFCs.However, these materials are vulnerable to carbon corrosion.The issues with the Pt catalyst, along with the weak interaction between the carbon support and the catalytic metal, do not assure consistent performance during long-term operations.Carbon corrosion in harsh environments causes vacancies in Pt NPs.Owing to the deterioration of the catalyst, the aggregation and detachment of Pt particles follow Ostwald aging and finally cause a decrease in the ECSA. [43]Strong metal-support interaction (SMSI) was suggested as the reason for the increase in the lattice constant of Pt.Several studies have reported the SMSI phenomenon. [44,45]In SMSI, the electrons of the transition metal atom of a metal oxide are transferred to another metal.The Pt metal that receives the electrons is filled with electrons in the d-band region; hence, the anti-bonding with oxygen required for the ORR performance of Pt increases.Antibonding with oxygen increases the reduction rate of oxygen and enhances its effect. [46]In this study, a carbon-metal oxide composite is designed to improve performance and durability.The composite adopts the concept of SMSI, [47] which contributes to electron density, structure, and protection through a metal oxide coating on carbon, thereby increasing its durability and catalytic activity.In the designed composite, the metal oxide layer is coated on a multi-walled carbon nanotube (MWCNT), and Pt is fixed, as displayed in Figure 2. Figure 2a shows the existing catalyst, and Figure 2b shows a conceptual diagram of a possible location for the designed protection effect.This concept suppresses the direct contact between water and carbon in the PEMFC environment and guarantees a life span in a specific environment.The material designed in this study consists of a transition metal oxide (SnO 2 ), which is a promising candidate for lowering the d-band center of Pt by transferring electrons to Pt. [48] In addition, the metal oxide exhibits stable characteristics in the acidic environment of a PEMFC.It can enhance the performance of Pt by effectively decomposing highly reactive oxygen species during the ORR and imparting strong corrosion resistance to the catalyst support.

Results and Discussion
In the experimental method, the temperature was set to 160 °C to facilitate the spontaneous formation of solid Pt particles from the dissolved Pt in the solution, lowering the Gibbs free energy to a negative value, which enables particle generation. [49]In the thermal synthesis of Pt-SnO 2 /MWCNT, the supported SnO 2 / MWCNT acts as a nucleation site for Pt synthesis, as depicted in Figure 3a.MWCNTs provide a suitable platform for the immobilization of Sn can be immobilized. [50]Additionally, the growth of Pt nuclei on SnO 2 -coated MWCNTs results in a larger surface area and a more extensive site for the growth of nucleation, primarily attributed to the presence of oxygen defects.Metal oxides are crucial in stabilizing low-coordination Pt sites and nucleation growth sites for synthesis. [51]Additionally, they contribute to stabilizing small Pt particles during voltage cycling, promoting the dispersion of small metal particles and enhancing the durability of ORR catalysts. [52]The Pt ion in the liquid solution contacted both the SnO 2 and MWCNT.Conducting investigations focused on the nuclear growth process, with a specific emphasis on surface energy, to collect evidence supporting the prioritization of Pt particle growth.During the thermal synthesis process, the formation of Pt particles can be characterized by the nucleation equation, which is referred to as Equation ( 2)- (5).
where r represents the radius of the synthesized solid Pt particles, γ represents the difference in surface energy, G v denotes the Gibbs free energy per unit volume, k B is the Boltzmann constant, T is the temperature, and θ is contact angle.As the experiments were conducted at the same temperature and concentration with a fixed volume, these parameters were considered constant.Therefore, the solidification of Pt ions into the solid state follows an equation that relates the Gibbs free energy of solidification to the surface energy and size of the Pt particles.
Table 1 presents the surface energy values obtained for different materials in previous studies.The reported surface energies of Pt, SnO 2 , and MWCNT-COOH are 1.84, 1.14, and 0.04 J m À2 , respectively.In Equation ( 4), ΔG v has a negative value, as the transition from the liquid to the solid state involves a decrease in volume.The contact angle depends on the surface energy.Therefore, as the difference in surface energy decreases, resulting in a larger negative value for ΔG heterogenous_tot , the spontaneous reaction occurs. [53,54]This reaction is illustrated graphically in Figure 3b.Based on previously investigated surface energy values, the Pt particles preferentially solidified on the SnO 2 / MWCNT support surface, where the surface energy difference was lower.Consequently, during the solidification of Pt ions, it was primarily synthesized on the SnO 2 surface with a lower activation energy, resulting in smaller Pt particles.The preferential synthesis of Pt particles on SnO 2 contributes to the enhanced effect of the SMSI.The particle sizes of Pt/MWCNT and Pt-SnO 2 /MWCNT obtained from the morphology analysis revealed that the Pt particles in Pt-SnO 2 /MWCNT are more uniformly dispersed and exhibit smaller sizes.Figure 3c illustrates the principle of SMSI.In the experimental process of this study, SnO 2 was formed by annealing in air after Sn ions were combined on the MWCNT.Then, Pt particles were formed on SnO 2 .In the following process, an SMSI is formed between a transition-metal oxide composed of Pt and Sn.Pt particles synthesized on SnO 2 exhibited different characteristics compared to those of Pt catalysts synthesized on conventional MWCNTs.When the SMSI phenomenon occurs in the designed Pt-SnO 2 /MWCT composite, Pt on SnO 2 receives electrons from Sn.Hence, as Pt atoms receive additional electrons, the interatomic distance expands due to the growing electron abundance.Consequently, the bonding strength between the particles diminishes.In Figure 3, the increase in the lattice constant is caused by the transfer of electrons from Sn to Pt owing to the SMSI effect between Pt and SnO 2 .Moreover, Pt exhibits a local increase in the electron density.SnO 2 is an oxide with low electrical conductivity, which has extensive applications in stabilizing the aggregation and dissolution of Pt NPs and inhibiting the decay of ORR activity. [55]SnO 2 on the MWCNT surface induces a synergistic effect that changes the adsorption energy and electrical conductivity, affecting the fixation and dispersion of Pt atoms at defect sites.The decrease in the d-band center position is attributed to compressive strain, which leads to an increase in electron density around the Pt atom due to charge transfer from the support.This, in turn, causes the d-band center to shift downward. [56]e increased electron density shifts the d-band center negatively as electrons fill the Pt d-band.A negatively shifted Pt d-band center indicates that the overall d-band energy of Pt was negatively shifted.As shown in Figure 3c, the lower d-band energy of Pt widens the d-band energy region lower than the Fermi level.This widening of the d-band energy region resulted in the same behavior of the Pt antibonding region near the Fermi level, and the Pt antibonding electrons were filled.In conclusion, with the activation of the antibonding region, the slow kinetic reaction rate during the ORR can be improved.
The X-ray diffraction (XRD) patterns representing the components and phases of the Pt-SnO 2 /MWCNT, Pt/MWCNTs, and Pt/C 20% are shown in Figure 4. Pt-SnO 2 /MWCNT, Pt/MWCNT, and Pt/C 20% consist of carbon as a support, which is confirmed by the common carbon diffraction peak at 2θ = 25.92°.The peaks corresponding to the Pt(111), Pt(200), and Pt(220) planes of Pt particles matched the XRD patterns of the Pt/MWCNT and Pt/C 20% catalysts.In the case of Pt-SnO 2 /MWCNT, the peaks of Pt(111), Pt(200), and Pt(220) shifted in the negative direction.It was confirmed that the peak shift increased from a high to a low angle.This explains why the properties of Pt particles follow Bragg's law (nλ = 2dsinθ). [57]here n is the diffraction order, λ is the wavelength, d is the lattice constant, and θ is the reflection angle.When Pt particles are bonded to the surface of SnO 2 , they are affected by the lattice constant, which is different from a general catalyst.Theoretically, Pt particles whose lattice constant is further increased by the influence of SnO 2 exhibit a peak shift toward a lower value of θ.MWCNT-COOH 0.04 [77]  This shift maintains the lattice constant value as the d value increases.In addition, because the value of the lattice constant of Pt is smaller at high angles, even a small change in its value can lead to a larger peak shift in θ.As evidenced by prior studies, it has been established that the predominant growth of the (211) plane in SnO 2 occurs during thermal processing at temperatures surpassing 500 °C.The expression [58][59][60] is currently undergoing evaluation.Therefore, the experimental results provide confirmation that the heat treatment process took place through the preferential growth of the (211) plane of SnO 2 .The values derived from Equation ( 2) are presented in Table S2, Supporting Information.As depicted in the transmission electron microscope (TEM) image in Figure 5, the observed trend of particle size for Pt/C 20%, Pt-SnO 2 / MWCNT, and Pt/MWCNT is consistent.
Experimental considerations for the synthesized catalysts are presented in Figure 5.In the Pt/MWCNT shown in Figure 5c, it is confirmed that Pt particles were synthesized on the surfaces of the MWCNT.In Figure 5d, the HRTEM image of Pt/MWCNT shows a lattice constant of 2.260 Å, which is the interplanar distance of Pt(111) and 1.960 Å for Pt(200).This aligned well with the reference lattice constant of Pt. Figure 5e shows a TEM image of commercial Pt/C 20%.In Figure 5f, the Pt particles of 20% Pt/C are observed on the Pt (111) plane in the state of carbon.The observed Pt (111) has the same lattice constant as the Pt(111) plane of the Pt/MWCNT observed in Figure 5c, unlike in the case of the metal oxide.It has been reported that different crystal planes of Pt have different d-band center values.Pt (100), Pt (111), and Pt (110) have d-band center values of -2.90, -2.76, and -2.54 eV. [61]As shown in Figure 3, Pt preferentially grows on SnO 2 because of the difference in surface energy.The grown Pt grains show differences in the TEM images shown in Figure 5. Pt particles grown on carbon show Pt (111) planes, whereas those grown on SnO 2 have (100) and ( 200) planes.Owing to the difference in surface energy, the preferentially grown Pt moves to the (200) plane and grows into a Pt particle with a lower d-band center value, which has a positive effect on ORR activity.Figure 5g and S1, Supporting Information, present a mapping image of the Pt-SnO 2 /MWCNT and an accompanying EDS data table.In the mapping image of Pt-SnO 2 / MWCNT, it was observed that the SnO 2 particles were relatively large, while the Pt particles were small and uniformly distributed on the surface of the MWCNT.This distribution pattern was confirmed in the overall mapping image.Specifically, through the examination of the mapping images of Pt and Sn, it was verified that each particle was uniformly present.Furthermore, the depiction of the synthesized Pt particles on the SnO 2 supports with larger particles suggests that it functions as a catalyst capable of enhancing the XRD low-angle shift and the TEM lattice constant. [62]igure 6a shows the binding-energy spectra of the Pt 4f sites of the synthesized Pt/MWCNT and Pt-SnO 2 /MWCNT catalysts.The binding energy of the Pt/MWCNTs has intense peaks at 71.68 and 75.08 eV.However, Pt-SnO 2 /MWCNT has an intensity peak at 71.48 and 74.88 eV due to a negative shift in the binding energy compared to that of Pt/MWCNT, and a negative peak shift occurs in the binding energy by 0.2 eV.The lower binding energy in the X-ray photoelectron spectroscopy (XPS) spectra lowered the binding energy between the Pt atoms.The theoretical and experimental values coincide with the phenomenon in which the lattice constant increases owing to the local increase in the electron density of Pt due to the SMSI effect, and the binding energy of Pt decreases as the interatomic distance widens.Figure 6b shows the Sn 3 d XPS spectra of Sn 2þ and Sn 4þ , which exist as SnO 2 and show high intensity.SnO 2 /MWCNT had intense peaks at 486.58 and 494.98 eV binding energies.However, Pt-SnO 2 /MWCNTs have intensity peaks at 487.18 and 495.58 eV, where a positive shift occurred compared with that of SnO 2 /MWCNT.Figure 6c   substantiates the significant electronic interaction observed in the positive shift results of the Pt-SnO 2 /MWCNT peaks, as compared to the O 2 and O 3 peaks of SnO 2 /MWCNT.Overall, the presence of electrochemically active Pt NPs generates an additional surface charge on the surface. [63]This is in contrast to the negative shift of the Pt 4f XPS spectra shown in Figure 6a.Pt gains electrons, and Sn loses electrons, leading to mutually consistent results.The XPS profiles of Pt-SnO 2 /MWCNT and Pt/MWCNT, measured in the valence band area to obtain the d-band center, are shown in Figure 6d.The XPS spectra of the valence band area showed changes depending on the catalyst support.Other studies have also reported that the XPS spectra change for each catalyst in the valence band area according to the change in the support. [64]he energy of the Fermi level corresponds to the 0 eV point.Pt/MWCNT has an intense peak at a binding energy of 2-3 eV close to the Fermi level when viewing XPS spectra in the valence band area.In contrast, Pt-SnO 2 /MWCNT has an intense peak at a binding energy of 4-5 eV farther from the Fermi level than that of Pt/MWCNT.As the binding energy measured in the valence-band region moved away from the Fermi level, the d-band center of Pt decreased.Therefore, using the binding energy of the value obtained from the center of gravity of the graph through the XPS spectra of the valenceband region, the value of the d-band center can be obtained according to Equation ( 6).The formula is created from the equation for determining the center of gravity of a general Figure 6d, [65] and the value of the x-axis obtained through this coincides with the value of the d-band center.catalysts, suggesting that the ORR catalytic performance can be improved. [66]igure 7a-c and Table S1, Supporting Information, show the curves for Pt-SnO 2 /MWCNT, Pt/MWCNT, and Pt/C 20% measured using a three-electrode system before and after applying the accelerated stress test (AST) protocol.The electrode scan rate was 20 mVs À1 , and the voltage range was 0-1.4 V (vs.reversible hydrogen electrode (RHE)).As a result of analyzing the current density peak before and after applying the AST protocol, Pt-SnO 2 /MWCNT showed a current density of 1.489 mA cm À2 at around À0.08 V before AST.The catalyst exhibited a current density of 1.409 mA cm À2 after the AST.In the case of Pt/MWCNT, a current density peak of 2.167 mA cm À2 was observed before the AST protocol at about À0.24 V, and it decreased to a current density peak of 0.910 mA cm À2 after the AST protocol.For Pt/ C 20%, the initial current density peak before the AST was 2.886 mA cm À2 .After the AST protocol, the current density peak is 1.070 mA cm À2 .The initial current density peak was the highest for Pt/C (20%); however, after the AST, it was confirmed that the current density peak of the catalyst is higher for Pt-SnO 2 / MWCNT.Simultaneously, it was confirmed that the change in the current density peak was the least, and the current density peak changed by approximately 5.4%.Pt-SnO 2 /MWCNT exhibited a slight change in the current density; therefore, it was more resistant to the effects of carbon corrosion in the AST protocol.The ECSA of the Pt catalyst was calculated using the hydrogen desorption area of Pt obtained from the cyclic voltammetry (CV) curves, and the corresponding values are shown in Figure 7d.(Figure S2, Supporting Information).These values are reported to be similar to the ECSA values measured in a 0.5M sulfuric acid solution.The disparity in the electrolytes H 2 SO 4 and HClO 4 did not significantly impact the observed outcomes in the experiment.The provided data illustrates the presence of a semicircle in the impedance spectrum, which is indicative of the bulk surface charge transfer resistance (R ct ) of the active material at high frequencies.In the low-frequency range, the impedance spectrum shows the presence of the Warburg impedance, which represents the insertion kinetics of the ion active material channel.The point of intersection on the Z real axis signifies the presence of internal resistance in the electrolyte, also known as the series resistance R s of the electrolyte.The charge transfer resistance exhibits a trend indicating the most rapid reaction kinetics. [67]o gain insight into the reaction kinetics of Pt-SnO 2 /MWCNTs for the ORR, the electrochemical impedance spectroscopy (EIS) technique was employed for analysis.The resistance of R s is relatively higher when compared to Pt-SnO 2 /MWCNT.Figure 7e depicts the catalyst observed at three distinct electrodes.The Nyquist plot of EIS revealed that the Pt-SnO 2 /MWCNT, Pt/ C(20 wt%), and Pt/MWCNT catalysts exhibited R ct values of 9, 11.41, and 17.02 Ω, respectively.These results indicate the reaction kinetics and ORR characteristics, with lower R ct values suggesting enhanced performance.The value of ion diffusion at the solid-liquid interface is enhanced, and a slope with a significantly positive value suggests that ion diffusion in the solution is relatively rapid.Among the tested materials, Pt-SnO 2 / MWCNT exhibits the highest value.Figure 7f illustrates the EIS spectrum for each catalyst after AST at the three electrodes.In the case of Pt/MWCNT, it is evident that the reaction kinetics of ORR are diminished as a result of the applied stress.As a result, the Pt-SnO 2 /MWCNTs are believed to have superior catalytic performance and durability.
The synthesized Pt-SnO 2 /MWCNT, Pt/MWCNT, and commercial Pt/C 20% were tested in a 0.5 M H 2 SO 4 solution saturated with N 2 solution using rotating ring-disk electrode (RRDE) to confirm the ORR performance.The measurements were analyzed by linear sweep voltammetry (LSV) under rotation conditions of 0, 100, 400, 900, 1600, and 2500 rpm, and the results of the Pt-SnO 2 /MWCNT, Pt/MWCNT, and Pt/C 20% are shown in Figure 8a,c,e.As shown in Figure 8a,c,e, the lower graph shows the current density with respect to the catalyst on the carbon disk.The current density of Pt-SnO 2 / MWCNT was higher than those of Pt/MWCNT and Pt/C (20%).The value of the disk with a higher current density under the measured conditions after coating with the same amount of catalyst showed that the performance of the catalyst increased.In addition, with the increase in the rotational speed the mass transfer rate increased, smoothening the mass transfer.Subsequently, the limiting current density increased.In the upper parts of Figure 8a,c,e, the catalyst inside the disk applies a constant voltage to the platinum ring to catalyze the catalytic reaction and meets the platinum ring to generate current, generating hydrogen peroxide.Therefore, in the case of the upper graph, the lower the current density, the more favorable the reaction.The graph shows a relatively low current value for Pt-SnO 2 /MWCNT.Through the relationship between the actual disk current density and ring current density, the number of electrons (n) and H 2 O 2 yield (p) are defined by Equation 7 and 8, respectively, where i d is the current density of the disk and i r is the current density of the ring.When N c = 0.424, the theoretical value of the RRDE product was used as the collection efficiency.Water was generated when the catalytic reaction became normal.At this time, it is an ideal reaction in which the number of moving electrons is four, so the closer n = 4 is to the experimental value, the more ideal the catalytic reaction.Similarly, the H 2 O 2 yield (p) was calculated using Equation ( 8). [68] Figure 8b,d,f shows the ORR electron numbers for the catalysts: Pt-SnO 2 /MWCNT, Pt/MWCNT, and Pt/C 20%.The three catalysts exhibited values of 3.99, 3.97, and 3.97, respectively.H 2 O 2 yields were 1.02%, 2.42%, and 1.97%.The onset potentials of the Pt-SnO 2 /MWCNT, Pt/MWCNT, and Pt/C (20%) catalysts are 0.785, 0.707, and 0.784 V, respectively, and their halfpotentials are 0.639, 0.563, and 0.597 V, respectively.As a result, Pt-SnO 2 /MWCNT reacted more fully than Pt/MWCNT and Pt/C (20%).The reaction occurred first through a high onset potential and had a high half-potential and fast catalytic kinetics.A lower H 2 O 2 production value, which is a side reaction, indicates high catalyst efficiency.
A unit cell test was conducted using the MEA production.The current-voltage (I-V ) curves obtained before and after the AST protocol are shown in Figure 9a-c.In the order of Pt-SnO 2 /MWCNT, Pt/MWCNT, and Pt/C 20%, the maximum power densities before the AST protocol were 0.515, 0.392, and 0.523 W cm À2 , and the open-circuit voltages were 0.959, 0.916, and 0.957 V. Before the AST protocol, Pt/C 20% had a higher maximum power density value, but the OCV showed a higher value for Pt-SnO 2 /MWCNT.Pt particles, whose d-band center is shifted in the negative direction according to the d-band theory due to the SMSI effect, have been confirmed to improve ORR performance.Prior to the AST, OCV was observed to have the highest Pt-SnO 2 /MWCNT ratio owing to its increased catalytic activity.After the AST protocol, the maximum power densities were 0.366, 0.079, and 0.318 W cm À2 for Pt-SnO 2 /MWCNT, Pt/MWCNT, and Pt/C 20%, respectively, and the OCV was 0.95, 0.901, and 0.957 V.The OCV values before and after the AST protocol showed slight changes for each catalyst.However, the change in the powder density had a significant effect.Pt-SnO 2 /MWCNT, Pt/MWCNTs, and Pt/C 20% exhibited power density changes of 29.84%, 79.84%, and 39.20%, respectively, after the AST protocol.The change in power density was small and followed the order Pt-SnO 2 /MWCNT > Pt/C 20% > Pt/MWCNT.This confirms that the SnO 2 /MWCNT catalyst support can help secure the durability of fuel cells by suppressing the decrease in catalytic activity and the drop-off and aggregation of the Pt catalyst particles.The ECSA was measured every 0, 100, 500, 1000, 2000, 5000, and 10 000 cycles between the AST protocols of the unit cell test.Pt-SnO 2 /MWCNT, 76.71% for Pt/MWCNT, and 57.84% for 20% Pt/C.The initial ECSA values were for the Pt-SnO 2 / MWCNT, which showed the largest hydrogen desorption area.This seems to be due to the fact that the selectively generated Pt particles on SnO 2 are smaller than those generated on normal carbon, the dispersion is increased, and the catalyst that was least affected by the AST protocol was also Pt-SnO 2 /MWCNT, which means that SnO 2 effectively inhibits the carbon corrosion of the carbon support.Pt-SnO 2 /MWCNTs do not meet the 40% ECSA reduction goal of the Department of Energy (DOE).However, considering that the ECSA reduction was less than 20% of that of commercial Pt/C, Pt-SnO 2 /MWCNT positively affected the durability of the carbon support.As shown in Figure 9d, Pt/C and Pt-SnO 2 /MWCNT have a larger area with 20% Pt/C as the initial ECSA area, whereas Pt-SnO 2 /MWCNT has a wider ECSA area after 5000 cycles of the AST protocol.It has also been proven that the catalyst durability is good.As shown in Figure 7 d, the change in the ECSA area measured in the three-electrode system and the trend of the ECSA change measured in the MEA coincided, indicating that Pt-SnO 2 /MWCNT has strong durability in the AST protocol related to corrosion of the catalyst support.Figure 9d shows the specific activity at 0.6 V (vs.RHE) to see the catalyst activity in the fuel cell operating voltage range.Prior to the AST protocol, the specific activity at 0.6 V was 0.442 for Pt-SnO 2 /MWCNT, 0.2915 for Pt/MWCNT, and 0.5033 A cm À2 for Pt/C 20%.After the AST protocol, Pt-SnO 2 /MWCNT, Pt/MWCNT, and Pt/C 20% had 0.298, 0.04693, and 0.261 A cm À2 , respectively.The Pt-SnO 2 /MWCNT reduced the specific activity by 32.7%, which is suitable for a reduction of less than 40%, as suggested by the DOE.In contrast, the specific activity decreased to 83.9% for Pt/MWCNT and 48.2% for Pt/C 20%; therefore, the reduction in catalytic activity suggested by the DOE did not exceed 40%.According to the AST protocol, a potential change of 1.5 A cm À2 was investigated.Before the AST, Pt-SnO 2 /MWCNT, Pt/MWCNT, and Pt/C 20% had voltages of 0.3423, 0.2411, and 0.3437 V, respectively.After the AST protocol, the voltages were 0.1984, 0, and 0.1060 V.
In the DOE, the potential change at 1.  sufficient values due to the SMSI effect and the downshift of the d-band center due to the use of the SnO 2 /MWCNT catalyst support, and the reduction due to durability is small.Figure 8g shows a Nyquist plot derived from EIS spectra analyzed on MEA with catalyst-specific applications on unicell.The resistance values (R s ) exhibit a close similarity, with values of 66.17 and 70.48.The variation in Rs values across the three electrodes can be attributed to the enhancement of contact resistance between the diffusion medium separating the anode and cathode sides and the bipolar plate.The resistance (R ct ) values obtained from the EIS analysis were 115.77, 124.39, and 126.38 Ω for the Pt-SnO 2 /MWCNT, Pt/C (20 wt%), and Pt/MWCNT catalysts, respectively.These results indicate the exceptional dynamic characteristics of the ORR for these catalysts.The reported Rs value for Pt-SnO 2 /MWCNT is comparatively lower than that of other catalysts.The EIS results obtained from a unit cell employing a catalyst via the SMSI effect exhibit a good agreement with the I-V curve (Figure 9a-c).In the unit cell test, the cell test using the MEA showed that Pt-SnO 2 /MWCNT achieved some of the DOE conditions, and at the same time, the performance of the catalyst was preserved compared to the data for Pt/C 20% and Pt/ MWCNT of other comparative groups (Table 2).By comparing the power density value of the cell test studied in other recent studies with that of this study, it was confirmed that the power density is higher than that of other studies.
The shape of the catalyst before the TEM images confirmed the AST protocol, to detail the corrosion process of the electrochemical and cell data, as shown in Figure 10a-c.Figure 10a-c shows that the average Pt particle size of Pt-SnO 2 /MWCNT is 2.97 nm, and that of Pt/MWCNT and Pt/C 20% is 4.73 and 4.57 nm, respectively.The fact that Pt-SnO 2 /MWCNT has smaller Pt particles than Pt/C 20% and Pt/MWCNT indicates a better dispersion of Pt NPs in the synthesis.Previous studies also confirmed that Pt NPs are evenly dispersed in these metal oxides. [69]The shapes of the catalysts after the AST protocol in the three-electrode system are shown in Figure 10d-f.Pt-SnO 2 / MWCNT increased the particle size of Pt, Pt/MWCNTs, and Pt/C 20% to 3.34, 6.25, and 5.81 nm, respectively.The particle size change according to the ECSA change after the AST protocol is confirmed in Figure 7d.In the case of Pt-SnO 2 /MWCNT, which showed relatively little change in ECSA, the smallest change in the Pt particle size was confirmed among the analyzed catalysts.In addition, Pt catalysts were present uniformly, and it was confirmed that there was little aggregation of the particles.However, in the case of Pt/MWCNTs, the dropout of the particles was accelerated, and the number of Pt particles was significantly reduced compared to the TEM image before the AST protocol, which is the cause of the lowest ECSA value.The Pt particles in the initial 20% Pt/C catalyst are uniformly distributed.After the AST protocol, the Pt particles of 20% Pt/C significantly increased the number of large particles due to the aggregation of the catalyst, which increased the average particle size of Pt. Figure 10g-i shows the TEM images of the catalyst after the MEA AST protocol.In Figure 10d-f, the micrographs are similar to the images of catalysts observed after the AST protocol.In Pt-SnO 2 /MWCNT, Pt and SnO 2 particles were uniformly present on the MWCNT surface, and Pt/MWCNT and Pt/C 20% showed a large reduction in Pt particles because of long-term exposure to the AST protocol.This indicates that the ECSA results in Figure 7 and the trend of change in the I-V curve are consistent.

Conclusion
This study investigates using a SnO 2 /MWCNT catalyst support to enhance the durability and performance of Pt catalysts.SnO 2 was coated onto MWCNT using a sol-gel method, and Pt was added through hydrothermal synthesis.The presence of SnO 2 had two significant effects.First, Pt particles were preferentially synthesized on top of SnO 2 due to differences in surface energy, resulting in smaller Pt particles with improved crystallinity.Second, SnO 2 induced a negative shift in the d-band center of Pt, leading to a fast ORR, which is beneficial for catalytic performance.Third, SnO 2 prevented direct contact between carbon and water, reducing carbon corrosion in electrochemical reactions.This SnO 2 /MWCNT support had enhancing catalyst durability through the SMSI effect.The SMSI effect was more pronounced when Pt grew preferentially on SnO 2 .The study also observed an increase in the lattice constant of Pt particles, as evidenced by XRD and TEM results.Electron transfer from Sn to Pt increased electron density, resulting in a negative shift of the d-band center of Pt and shifts in XPS spectra.These shifts in binding energy contributed to improved catalytic performance.AST confirmed that Pt-SnO 2 /MWCNT had better durability than commercial catalysts and Pt/MWCNTs, demonstrating its ability to suppress carbon corrosion.The study also showed improved ORR performance, prevention of Pt aggregation and detachment, and enhanced durability of carbon supports.In summary, the presence of SnO 2 on MWCNT affects the orientation and d-band of Pt, leading to improved catalyst durability through the SMSI effect and influencing electrochemical properties.This study offers a straightforward method for d-band control of Pt using metal oxides, potentially increasing Pt efficiency and economic viability in future Pt catalyst designs.

Experimental Section
Preparation of Pt-SnO 2 /MWCNT: Tin (II) chloride anhydrous (99.9%), platinum (IV) chloride (99.99%), sodium hydroxide (98%), ethylene glycol (99.8%), hydrochloric acid (37%), isopropanol (70%), and Nafion 117 containing solution (%5% in a mixture) were purchased from Sigma Aldrich (Republic of Korea).For reproducibility, COOH-functionalized MWCNTs Pt/SnO 2 /C 0.25 [79]   PtCu 3 @PWO x 0.421 [80]   Pt 1 Ni 1 /C 0.280 [81]   Pt/CNT-Ti 3 C 2 T x 0.181 [82]   Pt-Ag/C 0.503 [83]   Pt-Ni/C 0.37 [84]  were purchased from US Research Nanomaterials, Inc. (USA).The 20% Pt on Vulcan XC-72 was purchased from Premetek Co. (USA), and Nafion211 was purchased from Fuel Cell Store (USA).All reagents were used without further purification.The SnO 2 /MWCNT catalyst support was prepared using the sol-gel method.Two grams of anhydrous tin (II) chloride (99.9%) were dissolved in water.Subsequently, 2 mL of hydrochloric acid was added to prevent insolubility due to the hydrolysis of tin chloride during the mixing reaction.After adding 160 mg of the COOH-functionalized acid-treated MWCNT to the aqueous tin chloride solution, the mixture was stirred for 24 h.The stirred SnO 2 /MWCNTs were washed and dried using deionized water and ethanol and then annealed at 500 °C for 2 h in an air conditioner to collect SnO 2 /MWCNT.Pt-SnO 2 /MWCNTs were synthesized using a recipe-made SnO 2 /MWCNT catalyst support. [70]The SnO 2 /MWCNT of 80 mg and PtCl 4 of 35 mg were put in two 10 mL of ethylene glycol solutions, and the mixture was then dispersed by sonication for 30 min.After the two solutions were completely dispersed, they were mixed and stirred for 1 h.During stirring, the mixture was titrated with a 0.1 M NaOH solution to pH 10.The stirred solution was placed in a Teflon-lined stainless-steel reactor and hydrothermally synthesized at 160 °C for 8 h.Then, the reactant Pt-SnO 2 /MWCNTs were centrifuged, washed with ethanol 2-3 times, and dried in a drying oven at 70 °C for one day, and the Pt-SnO 2 /MWCNTs were recovered.The synthesis of Pt/MWCNT was similar to that of Pt-SnO 2 /MWCNT.Fabrication of MEAs: The synthesized catalyst (Pt-SnO 2 /MWCNT) was slurried for manufacturing the MEA.The catalyst ink was prepared by stirring the catalyst, isopropanol, and Nafion solutions at a ratio of 1:5:3 for 24 h.A catalyst layer of thickness 60 μm was deposited on top of the mesoporous carbon on GDL using the tape-casting method. [71]After that, it was dried in a dry oven at 70 °C for 1 day.Based on Nafion 211, a Pt/C 60% GDE was placed in the anode, and a GDE coated with a synthetic catalyst was placed on the cathode.Pressure was applied with a hot press heated to 80 °C to fabricate an MEA.
Characterization of the Catalysts: The morphology and microstructure of the Pt-SnO 2 /MWCNT and Pt/MWCNT samples were observed using TEM (JEOL: ARM2000).The crystal structure of the synthesized sample was measured at a rate of 2 min À1 using XRD analysis (PANalytical, Empyrean) with Cu Ka (λ = 0.15406 nm).Particle size analysis was conducted using XRD analysis.The calculation was performed using the XRD data of the catalyst and the Scherrer formula (Equation ( 9)).The symbol λ denotes the X-ray wavelength, while θ represents the angle of incidence (in radians) at which the X-ray was emitted.According to the Scherrer formula, it could be observed that the particle size was inversely proportional to the full width at half maximum (FWHM). [72]¼ Kλ βcosθ β (9)   The synthesized Pt/MWCNT and Pt-SnO 2 /MWCNT catalysts were analyzed using XPS to obtain the spectral changes and valence bands of the catalysts and catalyst supports.The spectra were calibrated based on the carbon peak at 284.6 eV.The concentration of the synthesized catalysts was measured using inductively coupled plasma optical emission spectrometry (ICP-OES; Agilent, 5100).The Pt mass composition of the catalyst reached an approximate convergence of 20% (Table S1, Supporting Information).In the calculation of the ECSA, the Pt loading was determined by considering the specific gravity of the catalyst used, which was then reflected in the ECSA results.
Electrochemical Data: Electrochemical measurements of Pt-SnO 2 / MWCNT were performed using a three-electrode system (SP 300, Biologic).The electrode system contained glassy carbon (CHI104, CH Instruments, Inc.), a saturated calomel electrode (CHI150, CH Instruments, Inc.) as the reference electrode, and a Pt coil (E045A, redox.me)as the counter electrode.Carbon support corrosion test analysis confirmed carbon corrosion using the AST protocol.The AST protocol simulated a fuel cell in normal operating and fuel-shortage environments.The cycle was applied for 10 s at 1.058 V and 50 s at 1.658 V based on the RHE.This cycle was repeated 20 times. [73]In addition, cyclic voltammetry was conducted before and after the AST.The scan rate of the CV test was measured at 20 mV s À1 and was investigated and changes in the ECSA were investigated and analyzed.The ECSA was obtained using the hydrogen adsorption or desorption area of the Pt catalyst based on the amount of Pt loading analyzed by ICP-OES.Table S1, Supporting Information, shows the ICP-OES weight ratio of synthesized Pt-SnO 2 /MWCNT and Pt/MWCNT.The amount of charge (Q H ) was calculated by obtaining the hydrogen adsorption or desorption area using the scan rate.The ECSA per unit of Pt was obtained by dividing by the Pt loading (derived from Equation 10) The ECSA, calculated according to the number of Pt units, was used as a catalyst performance index for fuel cells. [74]SAðm 2 pt =g pt Þ ¼ EIS was conducted under open-circuit voltage conditions.During the experimental measurements, the sinusoidal amplitude modulation was adjusted to a constant value of 10 mV within the frequency range from 0.01 Hz to 100 kHz.The spectra were recorded in an acidic medium with a concentration of 0.5 M H 2 SO 4 .The measurements were conducted by starting at high frequencies and gradually scanning toward lower frequencies in a logarithmic manner.
Unit Cell Test: A unit cell test was used to analyze the catalyst durability and performance in detail.After assembling the unit cell using the manufactured MEA (5 cm 2 ), the steady-state polarization curve was performed under the conditions of 100% relative humidity and no back pressure at a cell temperature of 70 °C.The fuel cell was fueled using H 2 /O 2 at a flow rate of 0.3/0.33L min À1 .The AST protocol for the MEA was performed under H 2 /N 2 at a flow rate of 0.2/0.4L min À1 , which was the carbon support corrosion protocol adopted by DOE.Each cycle of AST included applications of 1.0-1.5 V for 1 s and 1.5-1.0V for 1 s, and the AST was measured for 100, 500, 1000, 2000, and 5000 cycles.After each cycle, the MEA was recovered using H 2 , N 2 , and O 2 .Subsequently, the MEA performance was measured at 0.1 V in ECSA and H 2 /O 2 conditions using SP300 under H 2 /N 2 conditions.To perform EIS analysis, a constant current alternating current (AC) signal, with a 10% direct current (DC) component, was utilized within the frequency range from 0.01 Hz to 100 kHz.This analysis was conducted under H 2 /N 2 conditions.

Figure 1 .
Figure 1.Schematic illustration of issues caused by carbon corrosion in catalysts.

Figure 3 .
Figure 3. a) Schematic illustration of the synthesis mechanism of Pt-SnO 2 /MWCNT, b) Gibbs free energy graph of Pt nucleation at MWCNT and SnO 2 , and c) schematic illustration of the Pt-SnO 2 /MWCNT strong metal-support interaction.
These factors result in the XRD patterns at 39.20°, 45.65°, and 65.5°instead of 39.76°, 46.24°, and 67.46°, which are the theoretical 2θ values of Pt (111), (200), and (220) crystal planes, respectively (ICDD no.00-004-0802).The confirmation of SnO 2 synthesis in the catalyst can be achieved by analyzing XRD spectral data and examining the SnO 2 profile.It has been confirmed that the prominent peaks of SnO 2 are observed at 26.5°, 33.8°, and 51.7°, indicating their alignment.Among the observed phenomena, it has been noted that the peak intensity of the (211) plane of SnO 2 at an angle of 51.7°exhibits a discrepancy compared to the peak intensity of the overall SnO 2 profile.The reason for this can be attributed to the heat treatment that was carried out during the synthesis process.
Figure 5a-g with images of Pt-SnO 2 / MWCNT, Pt/MWCNT, and Pt/C 20% observed using TEM and high-resolution TEM (HRTEM).As shown in Figure 5a, the TEM image of Pt-SnO 2 /MWCNT confirms that the Pt NPs on SnO 2 were uniformly formed.These NPs exist in the form of the Pt(200) planes, as confirmed by Figure 5b.Pt(200) formed on SnO 2 has a lattice constant of 1.988 Å.This has a constant lattice value greater than 1.960 Å, the constant lattice value of Pt (200) of the XRD reference.SnO 2 (200) bonded to Pt has a lattice constant of 2.316 Å, which is smaller than the reference lattice constant of SnO 2 (2.368 Å, ICDD No. 01-072-1147).As shown in Figure 4, the reason for the shift to a lower angle in the XRD pattern of Pt coincides with the expected result, and it is confirmed that SnO 2 increases the lattice constant of Pt.
displays the XPS spectra of O1 s for both SnO 2 /MWCNT and Pt-SnO 2 /MWCNT.The XPS spectra exhibit peaks corresponding to O1, O2, and O3.The O1 peak of SnO 2 /MWCNT exhibits a binding energy of 530.38 eV, whereas the O1 peak of Pt-SnO 2 / MWCNT demonstrates a positive shift to 531.08 eV.The presence of additional O peaks, namely O2 and O3 peaks, further
band center value calculated using this formula is shown in Figure6ed band center graph.The blue graph represents the d-band value of the Pt/MWCNTs, and the red graph represents that of the Pt-SnO 2 /MWCNT.The white line in the graph indicates the location of the d-band center value.The d-band center value of Pt/MWCNT obtained using Equation (8) was derived as -3.467 eV, and that of Pt-SnO 2 /MWCNT was derived as -4.506 eV.Compared with Pt/MWCNT, the d-band center value of Pt-SnO 2 /MWCNT was shifted by 1.039 eV.This follows the mechanism shown in Figure3cby filling electrons in the d-band of the Pt particle, which has a strong binding energy with oxygen owing to the transfer of electrons from the transition metal.According to this mechanism, the d-band center of Pt is downshifted, and the oxygen and bonding energies of Pt are weakened, resulting in a faster ORR than conventional Pt
The measured ECSA are shown in Figure 9d.In cycle 0, the Pt-SnO 2 /MWCNT, Pt/ MWCNT, and Pt/C 20% catalysts had ECSA of 134.95, 102.44, and 138.37 m 2 g Pt À1 .After that, at 500 cycles, ECSA were 134.71, 135.50, and 75.73.ECSA were 97.51, 71.46, and 33.49m 2 g Pt À1 at 2000 cycles, and 42.76, 34.70, and 13.12 m 2 g Pt À1 at 5000 cycles.The ECSA tended to decrease as the AST protocol progressed for all catalysts.Comparing the data before and after 5000 cycles of the AST protocol, the reduction in the ECSA was 47.28% for
5 A cm À2 aims to change within 30 mV.The Pt-SnO 2 /MWCNT, Pt/MWCNT, and Pt/C 20% catalysts exhibited changes of 143.9, 241.1, and 237.6 mV changes.Although it is less than the value of the potential change targeted by the DOE, it has a lower potential change than Pt/C 20% and Pt/MWCNT.The results show that the amount of current per unit area and voltage generated at 1.5 A cm À2 are

Table 2 .
Power densities of typical Pt-based catalysts.