Reactive Surface Explored by NAP-XPS: Why Ionic Conductors Are Promoters for Water Gas Shift Reaction

Reactive Surface Explored by NAP-XPS: Why Ionic Conductors Are Promoters for Water Gas Shift Reaction Nuria García-Moncada,*Anna Penkova, Miriam González-Castano, and JoséA. Odriozola Cite This: ACS Catal. 2024, 14, 14947−14957 Read Online ACCESS Metrics & More Article Recommendations * sı Supporting Information ABSTRACT: Near-ambient pressure X-ray photoelectron spectroscopy (NAP-XPS) experiments have been carried out in N2and N2−H2O atmospheres on a Pt-based catalyst physically mixed with an Eu-doped ZrO2ionic conductor as a function of temperature under realistic conditions of the water gas shift (WGS) reaction. This work aims to demonstrate the significant effect of having active H2O on the ionic conductor surface at reaction temperatures to provide it to Pt metal sites. The ionic conductor, Eu-doped zirconia matrix, presents defects (oxygen vacancies, Ov) that allows upon H2O dissociation the formation of a hydrogen-bonded molecular water layer favoring diffusion through a Grotthuss mechanism below 300 °C. In the presence of H2O, the Ovare occupied by hydroxyl species as observed in the Eu 4d spectra, which differentiate two types of Eu oxidation states. The Eu3+-to-Eu2+ atomic ratio increases with the occupancy of the Ovby hydroxyls. Moreover, while the Pt-based catalyst alone is unable to create Pt−OH bonds, the physical mixture of the Pt-based catalyst and the ionic conductor allows the formation of Pt−OH bonds from room temperature up to 300 °C. These data demonstrate that the increase in molecular water concentration on the ionic conductor surface up to 300 °C acts as a reservoir to provide water to the Pt surface, enhancing the catalyst performance in the WGS reaction, supporting the importance of the surface H2O concentration in the reaction kinetics. KEYWORDS: WGS, ionic conductor, water activation, NAP-XPS, Pt 4f, Eu 3d 1. INTRODUCTION The relevance of developing more efficient catalysts for water gas shift reaction (WGS, CO + H2O⇆CO2+ H2) is widely known. 1−3 For this purpose, clarification of the reaction mechanism and determination of the rate-determining step are decisive. It is commonly accepted that a redox mechanism for high-temperature water gas shift reaction (HT-WGSR) uses iron-based catalysts. However, the reaction mechanism, active intermediates, and the rate-determining steps at lower temperatures (LT-WGSR) using Cu-based catalysts or noblemetal-based catalysts are still controversial. 2−4 As can be concluded from many reported works, the reaction mechanism is sensitive to temperature, catalyst design, and feed composition. 1,3,5−10 In broad terms, noble metals (Pt, Au, Rh, etc.) have shown good activity in WGSR at low−middle temperatures, where the support of the catalyst plays a crucial role. 8,11−19 This support, such as CeO2, TiO2, ZrO2, or CeO2/ Al2O3, besides giving a higher surface and better active metal dispersion and stability, provides oxygen defects. These defects are involved in the different proposed mechanisms, usually associated with the adsorption of water and the formation of different intermediates like carboxylates or formates. 7,20−26 Moreover, the formation and/or desorption of these intermediates, or its reaction toward CO2formation, are also considered as slow steps but always promoted by higher coadsorption or coverage of H2O molecules. 20,21 Thus, the focus on improving the support and the interaction between the active metal and the support has been widely aimed. 3,27 Nevertheless, no matter the considered mechanism, H2O adsorption and dissociation are the initial steps and are reported as the rate-determining step. The activation of H2O molecules has been described to take place on the metal active site, on the metal−support interface, or on the support defects. 19,28−30 As described in the literature, the dominant mechanism can change for the same catalyst formulation depending on the temperature or particle−support interaction and size, but the importance of dynamic surfaces with high mobility and the availability of water molecules close to the Pt−CO activated sites to obtain high performance in the WGS reaction has been shown. 31,32 In the same sense, based on density functional theory (DFT) and kinetic studies on WGS Received: July 18, 2024 Revised: September 20, 2024 Accepted: September 20, 2024 Published: September 25, 2024 Research Articlepubs.acs.org/acscatalysis © 2024 The Authors. Published by American Chemical Society 14947 https://doi.org/10.1021/acscatal.4c04287 ACS Catal. 2024, 14, 14947−14957 This article is licensed under CC-BY 4.0 Downloaded via UNIV DE SEVILLA on March 6, 2025 at 16:12:25 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles. reaction with catalysts at mild temperatures, it is clear that the coverage of water molecules on the surface available and activated ready to interact with the adsorbed CO is fundamental for higher kinetics. 6,24,33−38 Thus, in the literature, it is always found that the kinetic order of the water concentration on the active surface significantly influences the reaction rate. Therefore, we proposed in a previous work by this research group a new formulation of the catalyst, including the addition of an ionic conductor to a typical Pt-based catalyst, in order to promote water activation and mobility. This proposal was studied by kinetic tests using an optimized ionic conductor based on Mo-doped solid solution of Eu-doped ZrO2(with the optimal amount of 5 mol % of Eu2O3in ZrO2matrix). 39 These kinetic tests were carried out in both powders and as structured catalysts using metal microchannel monoliths, considering thermal and diffusional effects. The following experimental kinetic equation was obtained: r kP AP D (1 ) 1iCO H O H 2 2 = + · where pressures are in Pa and the rate is in mol·s−1·gcat −1. The description of the equation and fitting is deeply described in the previous work. 39 Regarding the kinetic order of the H2O (β, fitted at 0.4 for powder sample and 0.6 for structured sample), it is in good agreement with other equations found in the literature, 24,36−38,40 highlighting again that the kinetic order of H2O concentration is significant for the reaction rate and can be decreased by supplying more activated water, in this case by the addition of an ionic conductor. We propose that the essential step to improve the kinetics goes through to provide and keep H2O on the surface in interaction with the active sites. Therefore, ensuring the presence of H2O available for the catalyst, in this approach, by the presence of an ionic conductor, we should improve the kinetics and the activity of the catalyst. Additionally, we studied by impedance spectroscopy 41 that the addition of an ionic conductor housing oxygen vacancies was capable of capturing H2O and promoting proton conductivity by Grotthuss mechanism. 42−44 The creation of oxygen vacancies enables the adsorption of OH species from H2O molecules, and the mobility of protons and vacancies allows the H2O diffusion toward the active sites in the catalyst. As is found in the bibliography, the initial increment of oxygen vacancies increases the conductivity up to an optimal quantity, from which the association of the oxygen vacancies inhibits its mobility. 43,45,46 Particularly, it was shown that a ZrO2doped with 5 mol % Eu2O3forms a solid solution mixed oxide with an optimal oxygen vacancies amount, showing the highest ionic conductivity in dry and wet conditions. 41 Moreover, it was observed that the conductivity was higher in wet conditions up to 300 °C, when the temperature achieved is too high to keep enough superficial H2O on the sample. This temperature range of effectiveness makes these ionic conductors perfectly suitable for WGS catalysts that are active at mild temperatures. In order to demonstrate the concept of the promoting effect of any ionic conductor added to any active catalyst in the WGS reaction, different pairs of ionic conductors and WGS catalysts were also tested. 47 The inactivity of the ionic conductors by themselves and the promoting effect only observed in the temperature range of the activity of each catalyst supported the theory of water conductivity promoter but not additional catalytic activity from these additives. Moreover, these materials were tested by diffusive reflectance infrared Fourier transform (DRIFTS) in operando conditions showing the importance of H2O molecules in the intermediates and WGS reaction mechanism, and they were tested in the WGS reaction simulating ideal and realistic conditions, observing a clear increase in the activity of the promoted catalyst by the ionic conductor. 48,49 This work aims to monitor the dry and wet surface by the X- ray photoelectron spectroscopy (XPS) technique of the optimized Eu-doped ZrO2ionic conductor mixed with Ptbased catalyst for WGS, thereby supporting the conductivity, catalytic, and kinetics results already tested. The mixed sample will be analyzed and then compared with its two components analyzed under the same conditions but separately (Eu-doped ZrO2ionic conductor and the Pt-based catalyst alone). Since the objective is to observe and demonstrate the presence of H2O in the catalyst favored by the ionic conductor addition in operando conditions, the dry and wet flows at different temperatures in the range of WGS reaction will be monitored and measured by a new advanced near-ambient pressure-XPS (NAP-XPS) system that allows to operate in flow under pressure and temperature while XPS analysis is carried out. 2. MATERIALS AND METHODS The synthesis of the mixed oxide, named ZrEu_5, aimed 5 mol % of Eu2O3in ZrO2matrix, is described in a previous work. 41 In addition, Pt/SiO2with 4 wt % of Pt loading was prepared by wetness impregnation of Si support (Aerosil200, Evonik) by using tetrammineplatinum(II) nitrate solutions ([Pt- (NH3)4(NO3)2], Johnson-Matthey, 2.645 wt % Pt). The solution was acidified with 1 M acetic acid at Pt/AcOH = 1:1.1 molar ratio. The obtained solid was dried overnight and calcined for 8 h at 350 °C (5 °C/min heating ramp). The study of the H2O adsorption and its effect on these materials was carried out by means of the NAP-XPS measurements at SGI XPS in the Research, Technology and Innovation Center of the University of Sevilla (CITIUS). This NAP-XPS system (designed and manufactured by SPECS) is equipped with a PHOIBOS 150 NAP hemispherical energy analyzer with one-dimensional (1D)-DLD detector, a μFOCUS monochromatic small spot X-ray source (Al Kα), a diode laser (976 nm) for sample heating, and an in situ NAP cell connected to a gas mixture system (Figure S.1) to allow measurements in operando conditions, that is, simulating atmosphere and temperature in realistic operating conditions. Three samples were tested: Pt/SiO2catalyst, ZrEu_5 ionic conductor, and the physical mixture of Pt/SiO2+ ZrEu_5 in a mass proportion of 1:5. All samples were pelletized and placed in the setup, inside the NAP cell, where they were submitted to ultrahigh vacuum (UHV) at room temperature for a day and then heated up to 350 °C for 1 h in UHV in order to clean or remove any adsorbed species like superficial water. Then, two tests were carried out on both samples: first, dry pure N2was flown through the NAP cell, keeping the pressure at 5 mbar. XPS was recorded at room temperature, 150, 250, and 350 °C. Analogous test was performed afterward with wet N2flow, passing the pure N2through a saturator previously to the entrance in the NAP cell. High-resolution XPS analysis was measured with a pass energy of 50 eV, steps of 0.05 eV, and dwell of 1 s, with a small-area lens opening (design of NAP cell limits the opening to 300 μm). ACS Catalysis pubs.acs.org/acscatalysis Research Article https://doi.org/10.1021/acscatal.4c04287 ACS Catal. 2024, 14, 14947−14957 14948 3. RESULTS AND DISCUSSION 3.1. Pt/SiO2.The Pt-based catalyst was tested under dry and wet N2atmospheres in the NAP-XPS system in the temperature range of 25−350 °C. Sample charging during XPS measurements was corrected in all spectra using Si 2p peak, at 103.6 eV, since SiO2support provides a good intensity of the signal and does not change nor is affected during the tests. Figure 1 shows the XPS result of Pt 4f lines recorded at different temperatures during both dry and wet N2flows. It is noticeable that both conditions exhibit very similar spectra, meaning that the H2O molecules are not being adsorbed in these mild conditions. The most intense doublet peaks corresponding to 4f7/2 and 4f5/2 at 70.1 and 73.5 eV, respectively, for dry conditions, are ascribed to a reduced Pt. A slight shift to higher binding energies is observed in wet conditions; however, same peaks and contributions can be observed, indicating that H2O molecules are not directly interacting with Pt sites. 50,51 Figure 1. High-resolution Pt 4f XPS of the Pt/SiO2sample in operando conditions with a pressure in the NAP cell of 5 mbar of dry N2(left) and wet N2(right), at different temperatures. Figure 2. High-resolution Pt 4f XPS of the Pt/SiO2sample in operando conditions with a pressure in the NAP cell of 5 mbar of (a) dry N2(for dry conditions) and (b) wet N2(for wet conditions), at different temperatures. Deconvolution of the peaks in detail. ACS Catalysis pubs.acs.org/acscatalysis Research Article https://doi.org/10.1021/acscatal.4c04287 ACS Catal. 2024, 14, 14947−14957 14949 A deeper analysis of the Pt peaks by deconvolution shows a second less intense doublet corresponding to a more oxidized Pt site (Figure 2). This second contribution is placed ca. 1.5 eV higher than the most intense doublet, corresponding to an oxidized Pt specie. Thus, it is noticeable that even in wet conditions, the main phase is reduced Pt with a small contribution of Pt2+, both present in dry and wet conditions, and no evidence of Pt−OH is exhibited. Supporting these observations, any contribution from OH species cannot be distinguished during the analysis of the O 1s core level (Figure 3). The absence of a peak attributed to Pt− OH evidences a weak interaction in these mild conditions between H2O and Pt. The intense peak with an fwhm of ca. 2 eV corresponding to the SiO2support appears at 533 eV as expected, and it exhibits no change or shift during tests at any temperature. A peak associated with superficial molecular water can only be found at 536.8 eV during wet test, indicating the presence of water and the absence of adsorption. The shift of the peak ascribed to the molecular water toward higher binding energies (537.3 eV) at a low temperature (25 °C) may indicate more gas phase character, as is reported in the literature on TiO2materials submitted to H2O pressure treatments. 52,53 They observed different binding energies of O 1s peak associated with molecular H2O depending on the strength of interaction between water molecules and the support. Thus, a shift to higher binding energies is observed when the interaction is lower and the gas phase character of water molecules is higher, while lower binding energies indicate that the H2O molecules are on the surface implying stronger interaction. However, the slightly higher interaction with the support from 150 °C stays stable on increasing temperature, indicating an unchangeable weak interaction between superficial H2O molecules and the support. When the wet flow is stopped, the superficial H2O is easily removed at these mild temperatures. The increment in the intensities observed on Pt and O peaks also appears on the rest of the components (Si and C). It is due to the experimental conditions. Then, in order to compare if there is a real change, the superficial composition of each specie by integration of the peaks was carried out by CASA software. The atomic compositional analysis provides a stable amount of each Pt type as well as for the rest of the components. Thus, ca. 30−34 at % Si, 2.3−2.5 at % Pt, 50−60 at % O in dry conditions, 64 at % O in wet conditions, and 0− 10 at % C are obtained. All of the components are very stable, except C. The change in the carbon content is the major factor Figure 3. High-resolution O 1s XPS of Pt/SiO2sample in operando conditions with a pressure in the NAP cell of 5 mbar of dry N2(left) and wet N2(right), at different temperatures. Figure 4. High-resolution Eu 4d XPS of ZrEu_5 sample in operando conditions with a pressure in the NAP cell of 5 mbar of (a) dry N2(for dry conditions) and (b) wet N2(for wet conditions), at different temperatures. Deconvolution of the peaks in detail. ACS Catalysis pubs.acs.org/acscatalysis Research Article https://doi.org/10.1021/acscatal.4c04287 ACS Catal. 2024, 14, 14947−14957 14950 that proportionally affects the rest of the components. The analysis on the C 1s line allows us to distinguish two types of C (Figure S.2). It can be highlighted that the C signal ascribed to C−C bonds (ca. 285 eV) increases with temperature in dry conditions, which can be explained by C-cracking assisted by Pt active sites. The deconvolution of these peaks leads to a small contribution (not plotted for clarity) by carbonate species at ca. 288 eV. The intense C−C peak is partially transformed to carbonates under wet conditions (peak at 288 eV). Both C species are easily removed on temperature increase under wet conditions (Figure S.3). Thus, carbon is almost eliminated at 150 °C under wet conditions. The integration of these C peaks (Figure S.4) reflects the observation described above. The C content forming C−C bonds increases with temperature under dry conditions, while the carbonate content is constantly around 1 at %. Then, the introduction of H2O slightly decreases the C−C amount, but significantly increases carbonates from a low temperature (25 °C). This is easily shown as CO3/C−C peaks ratio (Figure S.4). This ratio is significantly higher under wet conditions at low temperatures. The quantification of both contribution (C−C and carbonates) from 150 °C became difficult and relatively inaccurate due to the already low signal, but the removal of C−C species is almost complete, only retaining a small amount of carbonates, achieving a maximum carbonate/ carbon ratio at 150 °C, from which C species are almost all removed. Therefore, undesired carbon is favored in dry conditions in the presence of Pt sites, but H2O definitely eliminates it with mild temperatures. 3.2. ZrEu_5. Concerning the analysis of the ionic conductor alone, Zr 3d5/2 at 182.3 eV was used to correct the charge effect, since the C 1s level partially overlaps with the Eu 4d line, and Zr exhibits a very stable doublet for both dry and wet experiments during the whole tested temperature range (Figure S.5). Thus, Zr 3d presents a stable doublet due to the spin−orbit effect with 2.4 eV of splitting in both dry and wet conditions over all temperatures. In the case of Eu 4d line (Figure 4), a more intense peak at ca. 135.8 eV with a spin− orbit splitting of 5.6 eV is shown. A deeper analysis of the peaks reveals a shoulder around 130 eV. This is due to the most intense Eu 4d5/2 peak ascribed as the Eu2+ contribution. Thus, a second pair of Eu 4d peaks were added, as expected. 54 The position of these small contributions attributed to Eu2+ is slightly higher than a pure Eu2+ specie. Some works on the study of these shake-down lines associated with Eu2+ establish the position of Eu 4d5/2 peak in the 132−125 eV range. 55,56 Thus, Kumar et al. 57 reported a peak at ca. 125 eV of a Fedoped Eu2O3, while Kim et al. 54 obtained two types of Eu2+ at 128.2 and 134 eV of a EuCl2sample. Mercier and co-workers 55 evaluated the binding energy of the main Eu3+ and its shakedown satellite ascribed to Eu2+ of different commercial Eu samples, and they concluded that different europium oxocompounds present different binding energies explained by differences in ionicity/covalence character affected by the electronegativity of the different atoms bonded in the oxocompound. Thus, they obtained the highest binding energy of Eu3+ for Eu(NO3)3but also the lowest binding energy of its correspondent Eu2+, that is, the highest energy gap between species Eu3+ and Eu2+ of 10.2 eV. Meanwhile, the lowest binding energy gap of 6.4 eV between species Eu3+ and Eu2+ was found for Eu2O3having in between species like Eu2(C2O4)3, Eu2(SO4)3, Eu2(CO3)3, and Eu(acac)3. In particular, the binding energies for commercial pure Eu2O3 of Eu3+ 4d5/2 and Eu2+ 4d5/2 are reported at 134.9 and 128.5 eV, respectively. Then, the sample of this work is in good agreement with the reported energy ranges. Moreover, considering that our ionic conductor is an O-deficient europium zirconia oxide, whose oxygen vacancies provide a lower electronegativity character, that is, less ionic and more covalent character with respect to the pure Eu oxide, it explains the slight increase of the binding energies in this sample. Concerning the corresponding O 1s line (Figure S.6), a very stable intense peak at 530.3 eV corresponding to the ZrO2 matrix is shown. A second smaller contribution around 532.6− 531.8 eV, slightly shifted depending on the temperature, is ascribed to OH bonds in the oxide matrix. There is not a qualitative significant difference between dry and wet conditions. In the case of the test under wet conditions, also superficial molecular H2O is detected at higher binding energies, in the range of 535−539 eV. As expected, the O 1s line peak of H2O molecules oscillates its binding energy depending on the temperature and its interaction with the support. Thus, a higher binding energy at 300 °C is shown when the water molecules have more gas phase character due to the high temperature, while the lowest energies for the superficial water molecules are observed at 150 and 250 °C, when the interaction with the Ovis optimal according to impedance spectroscopy studies. 41 In addition, C 1s was also measured (Figure S.7), where a second contribution around 288.5 eV is shown. This peak could be attributed to carbonates and Eu 4p. Quantification of the surface composition by the integration of the peaks shows that oxygen oscillates in the range of 32−43 at %, being higher when water is introduced; Zr and Eu keep their superficial contents at 30−36 at % Zr and 6−9 at % Eu, respectively; and the C content starts at ca. 23 at % decreasing with temperature until 11 at %, removed more easily under wet conditions. Therefore, the total superficial composition is quite stable, where only a slight increment of Zr, Eu, and O components takes place proportionally to the noticeable removal of carbon. However, a deeper analysis of the deconvoluted Eu peaks shows a meaningful change in Eu surroundings. In particular, it is highlighted that the ratio between the peaks ascribed to Eu2+ (Eu 4d5/2 and Eu 4d3/2 at 131.1 and 136.9 eV, respectively) and the most intense couple peaks corresponding to Eu3+ (Eu 4d5/2 and Eu 4d3/2 at 135.7 and 141.3 eV, respectively) decreases with temperature but with different profiles for wet than for dry conditions. As is shown in Figure 5, the partial amount of Eu2+ is significantly decreased in favor of Eu3+ at lower temperatures, from 150 °C, due to the filling of Ovby the presence of H2O. Then, for temperatures above 300 °C, both conditions, dry and wet, are indistinguishable. Simultaneously, the superficial amount of water also significantly decreases from 150 °C. This supports the conclusions deduced from previous impedance spectroscopy work, 41 that is, this solid solution based on Eu-doped zirconia is capable of hosting OH−/O−species, optimal from 150 to 250 °C, and generating an ionic conducting layer using water molecules up to 300 °C, when the superficial adsorbed molecular H2O is no longer enough for the conducting purpose. 3.3. Pt/SiO2+ ZrEu_5 (1:5). According to the previous analysis of the catalyst alone, Si 2p was used to correct the charge effect. For this mixture, C 1s (Figure S.8) is not completely removed during the experiment as it was observed above for the catalyst alone since the ionic conductor is also present. As described above for the ionic conductor alone, in ACS Catalysis pubs.acs.org/acscatalysis Research Article https://doi.org/10.1021/acscatal.4c04287 ACS Catal. 2024, 14, 14947−14957 14951 addition to the more intense peak C 1s, a second smaller complex peak appears associated with Eu 4p with some contribution of the carbonates. As described above, the ZrEu_5 sample shows interaction between Ovand H2O molecules from 150 °C up to 300 °C. On the contrary, Pt/SiO2catalyst exhibits no significant interaction with H2O molecules by itself. However, when the physical mixture of both materials as catalytic bed is tested in the same conditions, significant changes are observed in Pt 4f (Figure 6). During dry test (Figure 6a), high-resolution Pt XPS exhibited two Pt 4f7/2 peaks at ca. 70 eV, ascribed to reduced Pt, and 72 eV according to the oxidized Pt, as was observed above for the isolated sample. Both Pt sites present a spin− orbital splitting value of 3.35 eV between 4f7/2 and 4f5/2 lines and area ratio of 4:3 typical of f spin doublets. A negligible shift lower than 0.6 eV toward lower binding energy can be appreciated upon increasing temperature, indicating good stability of the Pt catalyst during the dry test. Nevertheless, when H2O is added to the mixture, the Pt species are both oxidized (Figure 6b). That is, the main contribution of oxidized Pt is still present above 72 eV, but now a smaller pair of peaks at higher binding energies are observed, while the absence of reduced Pt is highlighted. This new Pt specie exhibits its Pt 4f7/2 peak at ca. 74 eV, corresponding to an oxidized Pt directly interacting with the hydroxyl groups of the activated water. This means that the presence of the ionic conductor joined to the Pt catalyst favors the formation of oxidized Pt and OH bonds, contrary to the completely reduced Pt when the catalyst was tested alone. This important observation supports our proposed synergism between the ionic conductor and the catalyst: the availability of superficial activated water molecules in the ionic conductor in tight contact with the catalyst, close enough and appropriate activated energy to allow the catalyst to create bonds between its active site (Pt) and the reactant (activated OH groups spread over the surrounding surface). Since it is a physical mixture with 5 times more amount of ionic conductor than catalyst, this formulation ensures enough surface contact between the ionic conductor (where the H2O activates and moves superficially by Grotthuss mechanism) 42−44 and the catalyst, offering thereby available activated OH species to the catalytic surface. Moreover, this important feature is kept during the whole temperature range, according to WGS reaction operating at mild temperatures. Thus, the presence of the ionic conductor not only promotes the access of the H2O to the Pt but also keeps it on increasing temperature in the range of interest. In the literature, a similar approach by quasi Figure 5. Superficial atomic ratio corresponding to Eu2+/Eu3+ species (left axis) and superficial atomic % of O corresponding to superficial molecular H2O (right axis) of ZrEu_5 sample, calculated by NAPXPS in dry and wet conditions on increasing temperature from RT up to 350 °C. Figure 6. High-resolution Pt 4f XPS of the mixed sample ZrEu_5 + Pt/SiO2in operando conditions with a pressure in the NAP cell of 5 mbar of (a) dry N2(for dry conditions) and (b) wet N2(for wet conditions), at different temperatures. Deconvolution of the peaks in detail. ACS Catalysis pubs.acs.org/acscatalysis Research Article https://doi.org/10.1021/acscatal.4c04287 ACS Catal. 2024, 14, 14947−14957 14952 in situ XPS experiments on Pt/Ce0.75Zr0.25O2for WGS reaction can be found. 58 In that study, Gorlova and co-workers submitted the catalyst to different feed compositions for WGS reaction and compared the catalytic results with the analysis of the catalyst surface by XPS. Despite this quasi in situ XPS not being able to be performed under flow and water being not present, they tested the catalyst under CO atmosphere at RT, 150, and 350 °C, as well as the H2+ CO2mixture at 300 °C. They concluded that Pt0was the main specie and Pt2+ was always small and constant (around 12− 13%), in good agreement with our results. Moreover, the main change observed on Pt0was transformed into a Ptδ+, at 71.4 eV, with lower oxidation state than Pt2+. In addition, they also observed better reducibility of the ceria support (higher Ce3+ vs Ce4+, and increase of oxygen vacancies) due to the doping by Zr. They concluded that oxidized Pt can be the source of the highly reactive oxygen by comparing the catalytic results with XPS data. In this work by operando XPS experiments, we can certify that the ionic conductor allows the adsorbed H2O to form the active Pt−OH sites, consequently explaining the improvement of the performance and favoring the kinetics of WGS reaction. It is also notable that, for the highest temperatures (above 250 °C), there is an appearance of a new pair of reduced Pt species. This is in good agreement with the previous observation on the ZrEu_5 sample, where the maximum interaction with H2O caused by the ionic conductor is about 150−250 °C. For higher temperatures, the superficial H2O presents higher gas character and less interaction/concentration on the surface, providing reduced Pt sites not interacting with H2O. This behavior also supports the results of conductivity and activity studied in previous works for WGS reaction with physical mixtures of ionic conductor and Ptbased catalyst. 41,47,48 Moreover, the effect of the catalyst is maximal around this temperature in all cases. Thus, for example, we have reported catalytic results on Pt- or Cu-based catalysts mixed with different ionic conductors compared to the bare catalyst alone for the WGS reaction. 47 The formulation using the Eu-doped zirconia as ionic conductor showed an increment of the CO consumption of ca. 10% at 230 °C. Moreover, the addition of Mo to the ionic conductor provided an increment of the CO conversion in realistic WGS conditions (feed 9% CO, 11% CO2, 30% H2O, 50% H2) from 250 °C, achieving CO conversions over 55% with the mixture at 350 °C instead of the 35% obtained with the catalyst alone. 48 Additionally, Figure 7 represents the superficial atomic composition of Pt species in dry and wet conditions on increasing temperature. As discussed above, there is not any significant change in total Pt composition; however, the contribution of every Pt type is significantly different. In wet conditions, the reduced Pt disappears, showing instead two types of oxidized Pt. Thus, from the lowest temperature, new oxidized Pt associated with hydroxyl groups appear and Pt2+ is significantly increased with respect to dry conditions. Nevertheless, from 250 °C in wet conditions, where the gas character of the superficial H2O decreases its adsorption, reduced Pt sites reappear, decreasing the oxidized Pt. The analysis of the Eu 4d lines corresponding to the Pt/SiO2 + ZrEu_5 mixed sample corroborates the H2O interaction discussed above. As plotted in Figure 8a for dry conditions, an intense pair of peaks at ca. 135 and 141.6 eV corresponding to doublet Eu 4d5/2 and Eu 4d3/2 of Eu3+ are noticeable, as obtained for the ZrEu_5 sample alone in dry conditions. Similarly, a second small contribution ascribed to Eu2+ is found, with 4d5/2 centered at ca. 131 eV and again a spin−orbit splitting of 5.6−5.8 eV, as expected. Both Eu peaks are slightly shifted to lower binding energies on increasing temperature. This shift is probably due to a partial deformation of the Eudoped ZrO2structure. Analyzing O 1s lines and Zr 3d peaks (Figure S.9), the same shift toward lower binding energies is observed on increasing temperature, in both dry and wet conditions. In Figure S.9a, corresponding to the O 1s energy region during the test in dry conditions, two independent peaks at 533.0 and 530.3 eV are exhibited, ascribed to SiO2and Eu-doped zirconia, respectively. 59 As expected, since Si support does not change during the whole test and Si 2p is again taken as reference, the oxygen from SiO2does not change regardless of the temperature. On the contrary, a shift toward lower binding energies of O 1s ascribed to the ionic conductor is appreciated upon increasing temperature, becoming significant from 250 °C. Moreover, at the highest temperature, this shift evolves into two peaks, one at the original energy and a second peak at a clearly shifted binding energy. Simultaneously, this shift is also shown in the Zr 3d (Figure S.9b). Thus, the binding energy observed for Zr 3d5/2 at 182.3 eV at RT shifts toward 181.5 eV at 350 °C, with a constant spin−orbit splitting of 2.4 eV for all temperatures. This behavior of O and Zr atoms, only observed in the mixture, indicates a partial phase transition of the Eu-doped zirconia solid solution at higher temperatures in contact with SiO2. According to the previous characterization, 41 Eu-doped ZrO2presents a defectpyrochlore structure, with octacoordinated Zr, whose Zr−O bonds are in the order of 2.1 Å. Meanwhile, the formation of ZrSiO2is reported to keep the stable tetrahedra organization of SiO2, including octacoordinated Zr with Zr−O bonds in the range of 2.1−2.4 Å, 60 but affecting the ionic/covalent character of the Zr−O bonds. 59 Thus, this partial formation of mixed Eudoped ZrSiO2slightly shifts the binding energies of Zr 3d and O 1s toward lower values, indicating more ionic character in the formation of the ZrSiO2phase (Figure S.9). When the wet conditions are tested, the same shift is observed again in Eu (Figure 8b), O (Figure S.9c), and Zr (Figure S.9d) peaks. However, when the H2O is added, an extra pair of peaks are observed from 250 °C in the three Figure 7. Superficial atomic composition of Pt 4f species of the ZrEu_5 + Pt/SiO2mixed sample, calculated by NAP-XPS in dry and wet conditions on increasing temperature from RT up to 350 °C. ACS Catalysis pubs.acs.org/acscatalysis Research Article https://doi.org/10.1021/acscatal.4c04287 ACS Catal. 2024, 14, 14947−14957 14953 species (Eu, O, Zr). In fact, these “new” peaks in all of the spectra are observed only from 250 °C, when the water interaction changes, according to previous ZrEu_5 and Pt analysis. Moreover, the position of these pairs of peaks in Zr and O lines, both at binding energies close to those obtained originally for the ZrEu_5 sample alone, may indicate that adsorption of OH−/O−species in the Ovpartially inhibits the formation of ZrSiO2. This behavior also supports that the interaction with H2O is promoted by the ionic conductor and it is more significant around 250 °C. In addition, it is remarkable that in the case of Eu, which is surrounded by Ov, this new pair of peaks appear at a higher binding energy (blue line, Figure 8b), corresponding to the filling of Ovby water adsorption, which provides a more oxidized character to the Eu sites. Thus, the superficial amount of Eu in this Pt/SiO2+ EuZr_5 mixed sample, obtained by integration of the peaks, exhibits a similar curve to that shown in the ZrEu_5 sample alone. In Figure 9, the representation of the Eu2+/Eu3+ ratio gives again a minimum around 250 °C in both tests (dry and wet). It is remarkable that this minimum, that is, the maximal oxidation of the Eu sites, is more pronounced under wet conditions, indicating the partial occupation of oxygen vacancies by the activated water molecules. Again, above this temperature, both conditions are similar, and the interaction with H2O is lower due to a high enough temperature, supporting the previous discussion for the ionic conductor alone. Additionally, the integration of the O 1s peak (Figure 9) associated with molecular H2O (ca. 537−538 eV, Figure S.8c) to obtain its superficial composition shows a decrease from 11.1 at % at 25 °C to 1.8 at % at 350 °C, exhibiting again a decrease of superficial water with increasing temperature and supporting again the conclusions from impedance spectroscopy, where the promotional effect disappeared for temperatures above 300 °C due to the lack of superficial water. 41 Besides, the position of the peak ascribed to H2O molecules is significantly shifted to higher binding energies at 300 °C, indicating higher gas character, while it is the lowest at 150 °C, indicating a stronger interaction with the support, favoring the conducting model previously proposed. For temperatures over 250 °C, the H2O molecules are then proposed not to be significantly adsorbed as superficial molecules anymore but still partially filling the Ov, explaining the observation of reduced Pt and the Eu peaks at higher binding energy, previously discussed. Figure 8. High-resolution Eu 4d XPS of the mixed sample ZrEu_5 + Pt/SiO2in operando conditions with a pressure in the NAP cell of 5 mbar of (a) dry N2(for dry conditions) and (b) wet N2(for wet conditions), at different temperatures. Deconvolution of the peaks is shown in detail. Figure 9. Superficial atomic ratio corresponding to Eu2+/Eu3+ species (left axis) and superficial atomic % of O corresponding to superficial molecular H2O (right axis) of the ZrEu_5 + Pt/SiO2mixed sample, calculated by NAP-XPS in dry and wet conditions on increasing temperature from RT up to 350 °C. ACS Catalysis pubs.acs.org/acscatalysis Research Article https://doi.org/10.1021/acscatal.4c04287 ACS Catal. 2024, 14, 14947−14957 14954 4. CONCLUSIONS The NAP-XPS tests on the mixed sample created by the physical mixture of a Eu-doped ZrO2solid solution as ionic conductor and a Pt/SiO2as catalyst in dry and wet conditions have demonstrated the activation of the water molecules on the Pt sites and Ovof the mixed oxide. It has been observed that the ZrEu_5 ionic conductor exhibits an interaction with H2O molecules due to the filling of the oxygen vacancies, affecting the oxidated character of the Eu surroundings. This interaction is optimal at temperatures around 150−250 °C. From 250 °C, the superficial amount of water is significantly decreased. On the other hand, the Pt/SiO2catalyst does not show any meaningful interaction with H2O by itself at mild temperatures. However, when the catalyst is physically mixed with the ionic conductor, the latter promotes Pt−OH bond formation during the temperature range of interest, explaining the promotional effect obtained by catalytic tests in previous works. It is demonstrated that Pt needs the presence of an ionic conductor at low temperatures (below 300 °C) to have activated water and form Pt−OH bonds. Therefore, this work on NAP-XPS tests demonstrates the proposed model of a created conducting layer on the ionic conductor in wet conditions, operating below 300 °C that it does not show a catalytic effect by itself or alone but supplies activated water to the Pt catalyst promoting its activity through its rate-limiting step, that is, H2O activation, when they are placed together (physical mixture). Moreover, the NAP-XPS tests explain why a maximum promoting catalytic effect was obtained in the WGS reaction around 150−250 °C when water interaction is maximum. ■ASSOCIATED CONTENT * sı Supporting Information The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acscatal.4c04287. Image and details of the new NAP-XPS system employed in the experiments and complementary highresolution XPS data analysis of the tested samples (PDF) ■AUTHOR INFORMATION Corresponding Author Nuria García-Moncada −Servicios Generales de Investigación, XPS, Centro de Investigación, Tecnología e Innovación de la Universidad de Sevilla (CITIUS), 41012 Sevilla, Spain; orcid.org/0000-0003-0218-1135; Email: [email protected] Authors Anna Penkova −Departamento de Química Inorgánica e Instituto de Ciencia de Materiales de Sevilla�Centro Mixto US-CSIC, 41092 Sevilla, Spain; orcid.org/0000-0001- 5792-2362 Miriam González-Castano−Departamento de Química Inorgánica e Instituto de Ciencia de Materiales de Sevilla� Centro Mixto US-CSIC, 41092 Sevilla, Spain; orcid.org/ 0000-0003-2575-8398 José A. Odriozola −Departamento de Química Inorgánica e Instituto de Ciencia de Materiales de Sevilla�Centro Mixto US-CSIC, 41092 Sevilla, Spain; orcid.org/0000-0002- 8283-0459 Complete contact information is available at: https://pubs.acs.org/10.1021/acscatal.4c04287 Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. N.G.-M.: Conceptualization, materials preparation, data acquisition, data curation, formal analysis, writing�original draft, writing�review and editing; A.P.: Formal analysis, writing�review and editing; M.G.-C.: Materials preparation, writing�review and editing; J.A.O.: Conceptualization, formal analysis, writing�review and editing. Notes The authors declare no competing financial interest. ■ACKNOWLEDGMENTS The authors thank SMART FTS project (PID2021- 126876OB-I00, Plan Estatal 2021-2023, Ministerio de Ciencia e Innovación) and NICER BIOFUELS project (PLEC2021- 008086, Plan Estatal 2017-2020, Ministerio de Ciencia e Innovación) for funding, and Servicios Generales de Investigación−XPS in Centro de Investigación, Tecnología e Innovación de la Universidad de Sevilla (CITIUS) for the availability of the NAP-XPS system. ■REFERENCES (1) Ratnasamy, C.; Wagner, J. P. Water Gas Shift Catalysis. Catal. Rev. 2009,51 (3), 325−440. (2) Smirniotis, P.; Gunugunuri, K. Water Gas Shift Reaction: Research Developments and Applications; Elsevier, 2015. (3) Baraj, E.; Ciahotny, K.; Hlincík, T. The Water Gas Shift Reaction: Catalysts and Reaction Mechanism. Fuel 2021,288, No. 119817, DOI: 10.1016/j.fuel.2020.119817. (4) Rhodes, C.; Hutchings, G. J.; Ward, A. M. Water-Gas Shift Reaction: Finding the Mechanistic Boundary. Catal. Today 1995,23 (1), 43−58. (5) Huang, S.-C.; Lin, C.-H.; Wang, J.-H. Trends of Water Gas Shift Reaction on Close-Packed Transition Metal Surfaces. J. Phys. Chem. C 2010,114 (21), 9826−9834. (6) Kalamaras, C. M.; Dionysiou, D. D.; Efstathiou, A. M. Mechanistic Studies of the Water−Gas Shift Reaction over Pt/ CexZr1−XO2 Catalysts: The Effect of Pt Particle Size and Zr Dopant. ACS Catal. 2012,2(12), 2729−2742. (7) Meunier, F. C.; Tibiletti, D.; Goguet, A.; Shekhtman, S.; Hardacre, C.; Burch, R. On the Complexity of the Water-Gas Shift Reaction Mechanism over a Pt/CeO2 Catalyst: Effect of the Temperature on the Reactivity of Formate Surface Species Studied by Operando DRIFT during Isotopic Transient at Chemical Steady- State. Catal. Today 2007,126 (1−2), 143−147. (8) Jeong, D. W.; Potdar, H. S.; Shim, J. O.; Jang, W. J.; Roh, H. S. H2 Production from a Single Stage Water-Gas Shift Reaction over Pt/ CeO2, Pt/ZrO2, and Pt/Ce(1-x)Zr (x)O2 Catalysts. Int. J. Hydrogen Energy 2013,38 (11), 4502−4507. (9) Querino, P. S.; Bispo, J. R. C.; Rangel, M. D. C. The Effect of Cerium on the Properties of Pt/ZrO2 Catalysts in the WGSR. Catal. Today 2005,107−108, 920−925. (10) Ebrahimi, P.; Kumar, A.; Khraisheh, M. A Review of Recent Advances in Water-Gas Shift Catalysis for Hydrogen Production. Emergent Mater. 2020,3, 881−917. (11) Castano, M. G.; Reina, T. R.; Ivanova, S.; Centeno, M. A.; Odriozola, J. A. Pt vs Au in Water-Gas Shift Reaction. J. Catal. 2014, 314, 1−9. (12) Park, J. B.; Graciani, J.; Evans, J.; Stacchiola, D.; Senanayake, S. D.; Barrio, L.; Liu, P.; Sanz, J. Fdez.; Hrbek, J.; Rodriguez, J. A. Gold, Copper, and Platinum Nanoparticles Dispersed on CeOx/TiO2(110) Surfaces: High Water-Gas Shift Activity and the Nature of the Mixed- Metal Oxide at the Nanometer Level. J. Am. Chem. Soc. 2010,132 (1), 356−363. ACS Catalysis pubs.acs.org/acscatalysis Research Article https://doi.org/10.1021/acscatal.4c04287 ACS Catal. 2024, 14, 14947−14957 14955