Processing math: 100%

Advanced Search
Yi Tu, Xingwang Cheng, Luchao Huang, Dongling Zhang, Jun Hu, Qian Xu, Honghe Ding, Junfa Zhu. Enhanced Methanol Decomposition via Metal-Support Interaction on Ni/CeO2(111) Surface[J]. Chinese Journal of Chemical Physics . DOI: 10.1063/1674-0068/cjcp2409130
Citation: Yi Tu, Xingwang Cheng, Luchao Huang, Dongling Zhang, Jun Hu, Qian Xu, Honghe Ding, Junfa Zhu. Enhanced Methanol Decomposition via Metal-Support Interaction on Ni/CeO2(111) Surface[J]. Chinese Journal of Chemical Physics . DOI: 10.1063/1674-0068/cjcp2409130

Enhanced Methanol Decomposition via Metal-Support Interaction on Ni/CeO2(111) Surface

More Information
  • Corresponding author:

    Jun Hu, E-mail: hjruyi@ustc.edu.cn

    Junfa Zhu, jfzhu@ustc.edu.cn

  • These authors contributed equally

  • Received Date: September 27, 2024
  • Accepted Date: October 27, 2024
  • This study explores the adsorption and reaction of methanol on the CeO2(111) and Ni/CeO2(111) surfaces, highlighting the essential role of metal-support interaction in methanol decomposition by a synergistic approach encompassing synchrotron radiation photoemission spectroscopy, X-ray photoelectron spectroscopy, infrared reflection and absorption spectroscopy, and temperature-programmed desorption. Our findings reveal that Ni deposited on the CeO2(111) surface, followed by annealing to 700 K, leads to the formation of Ce-O-Ni mixed oxide as the dominant phase. The Ni2+ species facilitate the methoxy decomposition into CO and H2 within 300–430 K, with a small amount of formaldehyde also forming at the edge sites of ceria. Additionally, some methoxy adsorbed on the bare CeO2 surface migrates to the Ce-O-Ni mixed oxide, where they decompose into CO and H2 at 500–600 K, accompanied by a portion of the methoxy interacting with ceria to generate formaldehyde. Upon exposure to methanol at 500 K, the Ni2+ species are reduced to metallic Ni0, alongside the formation of coke and Ni3C, ultimately resulting in catalyst deactivation. However, reintroducing O2 reactivates these sites by oxidizing metallic Ni0 and Ni3C species. This study highlights the pivotal role of metal-support interaction in promoting oxygen transfer from ceria to Ni, thereby enhancing methoxy decomposition and significantly improving the performance of Ni-based catalysts for methanol decomposition into CO and H2.

  • Methanol, an indispensable industrial chemical, serves as C1 intermediate in the production of formaldehyde, aromatics, ethylene, methyl ethers and acetic acid, etc. [15]. Additionally, it presents a viable alternative to traditional fossil fuels due to its high efficiency, ease of transport and storage, greater eco-friendliness, and low exhaust emissions [69]. Ni supported on ceria emerges as an effective catalyst for methanol-associated reactions, including methanol steam reforming (CH3OH + H2O → CO2 + 3H2) [10, 11] and methanol synthesis (CO + 2H2→CH3OH; CO2 + 3H2→CH3OH + H2O; CH4 + O2→CH3OH) [1215]. The metal-support interaction in Ni/ceria catalyst has led to an intriguing phenomenon known as “reverse oxygen spillover”, where oxygen atoms from the ceria substrate migrate onto the Ni nanoparticles, which have a significant impact on the structural reconstruction of the catalyst and play a critical role in optimizing its catalytic performance [1618]. Therefore, it is imperative to gain a comprehensive understanding of the adsorption and reaction behavior of methanol on the Ni/CeO2 surface at the atomic and molecular level through the use of a model system, which is rare so far [10].

    Many previous studies focused on the investigation of the interaction between Ni nanoparticles with CeO2(111) thin films, emphasizing the crucial role of the Ni-CeO2 interaction in elucidating its catalytic performance [10, 12, 1922]. For example, X-ray photoelectron spectroscopy (XPS) studies have demonstrated that Ni remains in a metallic state after deposition on the CeO2(111) surface. Annealing at higher temperatures (500–800 K) leads to the partial reduction of the ceria substrate, with Ni2+ species existing as Ce1–xNixO2–y, driven by the migration of Ni into the substrate [20]. However, scanning tunnelling microscope (STM) studies have revealed that annealing at elevated temperatures results in the formation of larger, less dense particles, attributed to particle sintering [21, 22]. Based on these inconsistent results, we continued to study the interaction between Ni and CeO2(111) [19]. It was found that upon annealing the Ni particles on CeO2(111), sintering of small Ni particles into large particles occurred, accompanied by encapsulation of Ni particles by Ce-Ni-O mixed oxide where Ni remains in Ni2+ state.

    Regarding the adsorption and reaction of methanol on ceria-supported Ni model catalyst, there is few studies which can be found so far. However, the adsorption and reaction of methanol on ceria surface have been extensively studied over the past decades [2330]. Experimental findings indicate that methanol undergoes adsorption and dissociation, forming methoxy at 200 K, with oxygen vacancy (OV) enhancing both methanol adsorption and dissociation processes [27]. Previous studies have reported that methoxy can decompose to formaldehyde at 500–600 K on the CeO2(111) surface [23, 30]. Moreover, methanol decomposition on ceria surface is highly sensitive to oxygen vacancies, which promote methoxy decomposition, forming CO and H2 at around 650 K after complete dehydrogenation [23, 30].

    On the ceria-supported metal nanoparticle model catalysts, methanol adsorption and reaction on Pt/CeO2(111) [31], Pd/CeO2(111) [32], Rh/CeO2–x(111) [30], Ni/CeO2–x(111) [10], have been investigated. It has been demonstrated that the metal nanoparticles can significantly influence the methanol reactions that occurred on the ceria support. For the Ni/ceria surface, after annealing Ni deposition on CeO2(111) and CeO1.8(111) surfaces, using synchrotron radiation photoemission spectroscopy (SRPES), Senanayake et al. found that the Ni2+ is the main phase on the former, while the metallic Ni0 is predominant on the latter [10]. The fully oxidized Ni/CeO2 surface exhibits catalytic inertness towards methanol reaction. In contrast, on the Ni/CeO1.8 surface, the methanol decomposition results in the formation of surface coke and nickel carbide (Ni3C). Moreover, the presence of water can enhance the selectivity of CO2 production through the metal-support interaction. Although this study employing SRPES has provided significant insights into the methanol decomposition on the Ni/ceria surfaces, a deeper exploration into the adsorption states and methanol decomposition mechanisms by a combination of surface science techniques such as SRPES, infrared reflection and absorption spectroscopy (IRRAS) and temperature-programmed desorption (TPD) is imperative.

    In this study, we focus on the adsorption and decomposition of methanol on the Ni/CeO2(111) interface. The products formed during the decomposition process were monitored through TPD. The chemical species generated on the surface after annealing the sample were examined by SRPES, XPS, resonant photoemission spectroscopy (RPES), and IRRAS. Our results indicate that the formation of Ce-O-Ni mixed oxide due to the metal-support interaction upon annealing the Ni/CeO2(111) surface significantly enhances the methanol decomposition at low temperatures, resulting in the desorption of CO and H2 at around 300–430 K. Annealing the Ni/CeO2(111) model catalyst under methanol atmosphere leads to the surface deactivation toward the methanol decomposition due to the reduction of Ni2+ and the formation of coke and Ni3C on the surface. However, the induction of oxygen can reoxidize Ni3C and metallic Ni0 to Ni2+ species, thereby reactivating the model catalyst.

    SRPES and RPES measurements were performed at the Catalysis and Surface Science Endstation at the BL11U beamline in the National Synchrotron Radiation Laboratory, Hefei, China. This beamline is an undulator beamline offering soft X-rays with photon energy ranging from 20 eV to 600 eV and resolution (EE) better than 104 at 29 eV. Detailed descriptions of this beamline can be found in our previous studies [33, 34]. For all experiments, methanol (CH3OH) ( > 99.8% purity) and deuterated methanol (CD3OD) ( > 99.8% purity) were purified by several freeze-pump-thaw cycles before use. The SRPES experiments for methanol adsorption were conducted by exposing the sample to 5×10−7 mbar methanol for 30 min at 300 K, followed by heating the surface to a specific temperature. The photon energies were set to 600 eV for O 1s and 400 eV for C 1s. The resonant photoelectron spectra of the valence band were recorded at different photon energies, 121.4, 124.8, and 115.0 eV, corresponding to the resonant enhancements of Ce3+, Ce4+ ions, and off-resonance conditions, respectively. The determination of the n(Ce3+)/(n(Ce3+)+n(Ce4+)) ratio from RPES was described in our previous studies [33, 35]. The XP spectra were taken using Al Kα as the excited X-ray source. The photoelectrons were collected at an angle of 60° with respect to surface normal.

    IRRAS measurements were performed at the Photoemission Endstation at BL10B beamline in the National Synchrotron Radiation Laboratory, Hefei, China [34, 36]. This beamline, connecting with a bending magnet, offers soft X-rays with photon energy ranging from 100 to 1000 eV with a resolving power (EE) better than 1000. The endstation comprises five chambers: an analysis chamber, a preparation chamber, a load-lock chamber, a high-pressure reactor, and an IR chamber with a base pressure better than 2×10−10 mbar. An FTIR spectrometer (Bruker VERTEX 80v) and a liquid nitrogen-cooled mercury cadmium telluride (LNMCT) detector are connected to the IR chamber through the differentially pumped CaF2 windows. The measurements were conducted at a scan velocity of 20 kHz and a resolution of 4 cm−1.

    TPD experiments were carried out in a separate UHV chamber with a based pressure of 1×10−10 mbar, which has been described previously [34, 36]. Briefly. the chamber was equipped with an Auger spectrometer, a differential-pumped Hiden quadrupole mass spectrometer (QMS) for TPD, a four-grid optics for low energy electron diffraction (LEED), several evaporators including Ce and Ni, a gas doser, and a special manipulator with resistive heating where the single crystal sample was held by two Ta wires. The sample temperature was measured by a K-type thermocouple that was attached directly to the side of the sample. During TPD measurements, the mass spectrometer approached the sample surface about 1 mm so that desorption only from the sample surface could be detected by QMS. For the TPD experiments, the model catalyst surface was exposed to 0.03 L (1 L=1.33×10−6 mbar·s) of deuterated methanol (CD3OD) gas at 100 K, followed by programmed heating at a rate of 2 K/s.

    The Cu(111) single crystalline disk (Mateck, 8 mm diameter, 2 mm thickness) was cleaned using several cycles of Ar+ ion sputtering and annealing until no impurity could be detected by XPS, and LEED gave a sharp (1×1) pattern with low background. The 3–4 nm CeO2 thin films were prepared by using procedures described in detail elsewhere [36, 37]. Briefly speaking, the epitaxially grown ordered CeO2(111) thin films were produced by evaporating Ce in 2×10−7 mbar oxygen environment onto Cu(111) with a gradually increased substrate temperature to 800 K. To promote films ordering by forming large flat terraces, the films need to be post-annealed at 850 K for 10 min. Ni was vapor-deposited onto the CeO2 thin film at 300 K using the resistive evaporations at a constant rate of 0.1 ML/min. The metal coverage was estimated based on the attenuation of Cu 2p signal after the metal was directly deposited onto Cu(111). One monolayer (ML) of Ni is here defined as 1.86 × 1015 atoms/cm2 with respect to their atomic densities for the close-packed (111) plane.

    A fully oxidized CeO2(111) thin film was grown as the support for Ni deposition to investigate the influence of Ni coverage on its chemical state. FIG. 1 (a) and (b) display the LEED patterns of clean Cu(111) and 4 nm thick CeO2(111) thin films grown on Cu(111) surface, respectively. The clean Cu(111) surface exhibits a sharp p(1×1) LEED pattern. After the growth of 4 nm CeO2(111) thin films on Cu(111), the LEED pattern originating from the Cu substrate becomes obscured. Instead, an array of additional hexagonal spots emerges, corresponding to the (1.5×1.5) CeO2(111)/Cu(111) structure, which is consistent with previous results [37].

    Figure  1.  LEED patterns taken from (a) clean Cu(111), Ep = 65 eV; (b) 4 nm CeO2(111), Ep = 70 eV. (c) Ni 2p spectra of indicated Ni/CeO2 surfaces; black line: Ni deposition on CeO2(111) at 300 K, red line: samples annealed at 700 K. (d) The ratio of n(Ce3+)/(n(Ce3+)+n(Ce4+)) determined from RPES of Ni deposited on CeO2(111) surfaces upon annealing to 700 K.

    XPS and RPES were employed to investigate the chemical state of Ni upon its deposition on the CeO2(111) surface and the interaction of Ni with CeO2(111) thin films. This was achieved by monitoring the Ni 2p spectra (FIG. 1(c)) and the ratio of n(Ce3+)/(n(Ce3+)+n(Ce4+)) (FIG. 1(d)). In general, the results were consistent with previous findings [10, 19, 21]. The bottom black spectrum in FIG. 1(c) symbolizes the CeO2(111) thin films covered with 0.4 ML Ni at 300 K. The primary Ni 2p3/2 peak, with a binding energy of 852.8 eV, matches that of bulk metallic Ni0, according to previous studies [19, 35, 38]. In addition, a small shoulder peak located at 854.6 eV was also observed, indicating the existence of a small amount of Ni2+ on the surface. On the other hand, the increase in the ratio of n(Ce3+)/(n(Ce3+)+n(Ce4+)) to approximately 14.7% was observed, suggesting that the ceria is partially reduced due to electron transfer from the Ni nanoparticles to the ceria [19, 21, 35], which is in consistence with the appearance of small amount of Ni2+ on the surface. Subsequent annealing at 700 K results in the transformation of metallic Ni0 to Ni2+, as evidenced by the emergence of the peak at 854.6 eV at the expense of initial peak at 852.8 eV. Interestingly, the decrease in the ratio of n(Ce3+)/(n(Ce3+)+n(Ce4+)) after annealing indicates the reoxidation of ceria. A similar phenomenon was also observed in previous studies [20, 21], likely attributed to the diffusion of oxygen vacancies into the subsurface. At higher Ni coverage (0.6 ML), the Ni deposition on the CeO2(111) surface exhibited an obvious higher Ni0/Ni2+ ratio, which is evidenced by a stronger peak at 852.8 eV, while the peak at 854.6 eV remains almost constant. Upon annealing to 700 K, a significant decrease in the intensities of Ni 2p peaks was observed, which is attributed to the sintering of Ni nanoparticles [21, 22]. In addition, the peak at 854.6 eV increased, while the peak at 852.8 eV decreased, suggesting a higher proportion of Ni2+ species [19, 21]. The peak fitting of Ni 2p3/2 peak suggests that ~10% and ~25% of the 0.4 ML and 0.6 ML Ni on CeO2(111) surface after annealing to 700 K remain as metallic Ni0 state (FIG. S1 in Supplemengtary materials (SM)), respectively. Similar to the behavior of 0.4 ML Ni/CeO2 surface, the ceria thin films undergo oxidation during annealing, leading to a decrease in the ratio of n(Ce3+)/(n(Ce3+)+n(Ce4+)) to around 1%. In comparison to the Ni 2p3/2 spectra of 0.4 ML NiO/CeO2(111) (FIG. S2 in SM), it is noteworthy that the satellite peak at around 861 eV related to NiO was nearly undetectable on both annealed Ni/CeO2(111) surfaces, confirming that the predominance of Ni2+ species exists in a Ce-O-Ni mixed oxide with only a small amount of NiO [19, 21, 39]. The result is in line with our previous study that indicated simultaneous sintering of small Ni particles into larger ones and encapsulation by a mixed oxide Ce-O-Ni layer [19].

    In order to maintain the reproducibility and morphological stability of Ni/CeO2 model catalyst during methanol adsorption and decomposition, all the experiments were performed on the surface of 0.4 ML Ni/CeO2(111) model catalyst annealed at 700 K. The methanol decomposition process was first monitored by C 1s spectra after methanol adsorption at 300 K, followed by stepwise heating to 600 K. The results are shown in FIG. 2(a). Initially, on the pristine CeO2(111) surface, CH3OH adsorption at 300 K resulted in the appearance of two distinct peaks at 286.6 eV and 289.3 eV, corresponding to the formation of methoxy and formate species, respectively [24, 31]. The formation of formate species is likely due to the oxidation of partial methoxy at lower coordination sites [25]. Our previous STM studies have elucidated the surface morphology, indicating the formation of many small ceria islands on the 4 nm thick CeO2(111) thin films [37]. These islands can cause many lower coordination sites near the perimeter of these oxide island for methoxy oxidation. As the temperature rose to 400 K, the intensity of the peak at 286.6 eV decreased, suggesting the methoxy decomposition on the surface [27]. Further heating to 500 K significantly reduced the intensity of both peaks. Finally, at 600 K, all carbon species disappeared, leaving behind a carbon-free surface. Concurrently, in the O 1s spectra, an additional peak at 531.8 eV appeared after methanol adsorption at 300 K (FIG. S3 in SM). This can be attributed to methoxy (CH3O-) and/or hydroxyl (OH-) groups [23, 31, 32]. Furthermore, a high energy tail at 532.6 eV, indicative of formate species formation, was also observed [24, 31, 32]. Both of them vanished at around 500 K, in alignment with changes observed in the C 1s spectra.

    Figure  2.  C 1s spectra, =400 eV, of methanol adsorbed on (a) CeO2(111), (b) 0.4 ML Ni/CeO2(111) surfaces at 300 K and annealed to as-indicated temperatures. (c) Ni 2p spectra, =1486.6 eV, of methanol adsorption and decomposition on 0.4 ML Ni/CeO2(111) surface. (d) The ratio of n(Ce3+)/(n(Ce3+)+n(Ce4+)) determined from RPES of methanol absorption and decomposition on as-indicated surfaces.

    In a parallel series of SRPES experiments, methanol adsorption on the 0.4 ML Ni/CeO2(111) surface was examined (FIG. 2(b) and FIG. S3 in SM). The results are closely paralleled with those observed on the pristine CeO2(111) surface. Following methanol adsorption at 300 K, the C 1s (O 1s) spectra revealed peaks for methoxy and formate species at 286.7 eV (531.8 eV) and 289.3 eV (532.6 eV), respectively [24, 31, 32]. This was followed by a decrease in intensity of these peaks decreased upon heating, ultimately leading to their complete disappearance at 600 K. The comparative concentrations of the various surface species indicated that the presence of Ni enhances methoxy formation due to an increased number of oxygen vacancies on the Ni/CeO2(111) surface (FIG. S4 in SM) [27]. Furthermore, the Ni 2p spectra show no significant alteration in the Ni state throughout the methanol adsorption and decomposition process (FIG. 2(c)).

    The evolution of the n(Ce3+)/(n(Ce3+)+n(Ce4+)) ratio reveals distinctive trends for methanol adsorption and decomposition on the CeO2(111) and 0.4 ML Ni/CeO2(111) surfaces (FIG. 2(d)). Specifically, methanol adsorption on the pristine CeO2(111) surface at 300 K resulted in an initial increase of the ratio to approximately 3%. However, this ratio decreased during subsequent annealing. Heating the surface to 600 K caused the ratio to decline to a level that is similar to its original value. In contrast, methanol adsorption on the 0.4 ML Ni/CeO2(111) surface led to the ceria reduction, elevating the n(Ce3+)/(n(Ce3+)+n(Ce4+)) ratio to around 6%. This increased ratio remained stable even after annealing to 400 K, unlike the case on the pristine CeO2(111) surface. However, this ratio decreased upon annealing to 500–600 K, suggesting the reoxidation of ceria. The increase in Ce3+ concentration observed at 300 K following methanol adsorption on both the CeO2(111) and 0.4 ML Ni/CeO2(111) surfaces can be ascribed to dissociation of methanol molecules, generating hydroxyl and methoxy groups [24, 31]. During the annealing process, the changes in Ce3+ concentration can be attributed to two primary factors: the methoxy decomposition to hydroxyl groups, and the migration of Ce3+ into deeper layers. Among these, the latter plays a dominant role in annealing the CeO2(111) surface to 600 K, thereby contributing significantly to the reoxidation process. In contrast, the Ce3+ concentration remains stable on the 0.4 ML Ni/CeO2(111) surface upon annealing to 400 K, with only methoxy decomposition observed (FIG. 2(b) and FIG. S3 in SM). This indicates that methoxy decomposition can affect the Ce3+ concentration on the 0.4 ML Ni/CeO2(111) surface. Interestingly, a similar phenomenon occurs during methanol decomposition on the Pt/CeO2(111) surface due to the complete dehydrogenation of methoxy on Pt species [31]. So, we propose that the complete dehydrogenation of methoxy on Ni2+ species, followed by hydrogen spillover to ceria, leads to the formation of hydroxyl groups and the reduction of adjacent Ce4+ centers to Ce3+, thereby stabilizing the Ce3+ concentration when being annealed to 400 K. Our finding highlights the different reaction pathways of methoxy on the pristine CeO2(111) surface versus the 0.4 ML Ni/CeO2(111) surface. Additionally, the formation of Ce-O-Ni mixed oxide via metal-support interaction during pre-annealing process may play a critical role in methanol decomposition.

    To comprehensively understand the impact of Ce-O-Ni mixed oxide in Ni/CeO2 model catalyst on methanol decomposition, the products resulting from the CD3OD decomposition on the pristine CeO2(111) and 0.4 ML Ni/CeO2(111) surfaces were examined using TPD (FIG. 3). Here, we monitored the desorption of D2, water, CO, formaldehyde, CD3OD, and CO2 identified by mass-to-charge ratios of 4, 20, 28, 32, 34, and 44, respectively. Upon 0.03 L CD3OD adsorption on the pristine CeO2(111) surface, notable detected products were water, CD3OD, and formaldehyde, which coincide with previous results [23, 30]. Note that the signal at m/z of 32 is ascribed not only to formaldehyde but also to CD3OD cracking fragment. The relative intensities of CD3OD cracking fragments can be calibrated by integrating the desorption signal from multilayers of CD3OD, thereby correcting the formaldehyde signal [29]. The corrected formaldehyde signal showcases two significant desorption peaks at ~280 K and ~580 K (FIG. S5 in SM). The peak at ~280 K is attributed to methoxy decomposition at the edge sites of ceria, while the peak at ~580 K corresponds to methoxy decomposition at the terrace sites [29] and the reaction of formate species with hydroxyl groups [40, 41]. Water evolution exhibits a pronounced peak at 200 K. The physisorbed CD3OD multilayers desorb at around 140 K, followed by the evolution of chemisorbed, recombinative CD3OD from the surface at 200 K. Notably, no other products such as D2 or CO were observed from the fully oxidized CeO2(111) surface. It should be mentioned that previous investigations have found that methanol reaction on a reduced ceria surface at 650 K could produce H2 and CO [23, 26, 30].

    Figure  3.  Temperature-programmed desorption spectra for 0.03 L of deuterated methanol at 100 K and heated with a linear ramp of 2 K/s on (a) CeO2(111) and (b) 0.4 ML Ni/CeO2(111) surfaces.

    On the 0.4 ML Ni/CeO2(111) surface, upon exposure to 0.03 ML CD3OD at 100 K, the TPD spectra showed distinct difference as compared to those on the CeO2(111) surface. Except for the low-temperature desorption peaks of water and multilayer methanol, CO and D2 emerged as major reaction products. Broad desorption features of D2 within 300–430 K, as well as a desorption peak of CO at ~360 K were observed, which resemble those from methanol decomposition on the O-Ni(111) surface [42]. Notably, a minor quantity of formaldehyde was detected within the same temperature range, corresponding to methoxy decomposition at the edge sites of the ceria [29]. Furthermore, a notable broad peak of CO at ~590 K accompanied by a weak peak of D2 was observed. However, as the production of neither CO nor H2 at this temperature range is typically observed on either the CeO2–x(111) [23, 30] or O-Ni(111) surfaces [42]. Taking into account the extremely low Ce3+ concentration on the pre-annealed 0.4 ML Ni/CeO2 surface (FIG. 1(c) and FIG. 2(d)), the emergence of CO and D2 peaks at ~590 K cannot be attributed to the methoxy decomposition on the bare ceria surface, and it is likely that methoxy adsorbed on the bare ceria surface migrates to Ni2+ sites for decomposition. The weaker intensity of the D2 desorption peak can be attributed to the consumption of partial hydrogen atoms in reaction with formate species to produce formaldehyde [40, 41]. This hypothesis is supported by the formaldehyde desorption peak observed at 500–600 K, Additionally, methoxy decomposition at the terrace sites of ceria can produce formaldehyde within the same temperature range [29]. However, only methoxy species are observed in the corresponding with no formate species present at 500 K in the corresponding C 1s spectra (FIG. 2(b)). This can be attributed to different annealing processes used in the XPS experiments. The XP spectra were obtained after annealing the sample at the as indicated temperatures for one minute, during which the formate species were completely expended.

    To elucidate the adsorption and reaction mechanism of CH3OH on the 0.4 ML Ni/CeO2(111) surface, we conducted IRRAS experiments. These experiments involved exposing both the CeO2(111) and 0.4 ML Ni/CeO2(111) surfaces to methanol at 300 K, followed by annealing to specified temperatures (FIG. 4). Initially, methanol adsorption on the pristine CeO2(111) surface at 300 K resulted in three distinct peaks within the ν(CO) region at 1013, 1049, and 1103 cm−1, indicating different adsorption sites for methoxy species. Specifically, the peak at 1103 cm−1 is assigned with on-top-bonded methoxy species (type-I), while the peaks at 1049 and 1013 cm−1 correspond to bridge-bonded (type-II) and triply-bonded (type-III) methoxy species on the ceria surface, respectively [25]. Additionally, the significant peaks at 1356 and 1377 cm−1 were identified as the νs(OCO) mode of bidentate formate species displaying C2v symmetry, aligning with prior researches [43, 44]. Subsequent annealing at 400 K resulted in the disappearance of the peak for type-I methoxy species at 1103 cm−1. Conversely, the peaks for type-II and type-III methoxy species shifted to 1044, and 1011cm−1, respectively. Furthermore, the νs(OCO) peaks showed a slight increase at 1356 cm−1, with the 1377 cm−1 peak shifting to 1372 cm−1. By 500 K, all ν(CO) peaks vanished, leaving only the νs(OCO) peak for bidentate formate species. Finally, at 600 K, no discernible species was present.

    Figure  4.  IRRA spectra of methanol adsorbed on (a) CeO2(111), (b) 0.4 ML Ni/CeO2(111) surfaces at 300 K and annealed to as indicated temperatures.

    The IRRAS results from CH3OH adsorption on the 0.4 ML Ni/CeO2(111) surface at 300 K displayed the similar ν(CO) region characteristics to those observed on the pristine CeO2(111) surface. Specifically, the peaks at 1103 and 1024 cm−1 can be attributed to type-I and type-III methoxy species, respectively [25]. The observed blue-shift and higher intensity of the peak corresponding to type-III methoxy species can be attributed to a slight reduction of the ceria (FIG. 1(c)) [25]. Notably, the peak at 1065 cm−1 exhibited a larger full width at half maximum(FWHM) and a blue-shift compared to the peak of type-II methoxy species on the pristine CeO2(111) surface at 1049 cm−1 [25]. Previous research has identified the peak of methoxy on the NiO surface at 1060−1080 cm−1 [45]. Additionally, the peak of type-II methoxy species on the ceria surface has been observed at 1040−1060 cm−1 [25, 29]. Given these findings, it is suggested that the broad peak at 1065 cm−1 likely contains unresolved structures due to contributions from both bridging type-II methoxy species on the ceria surface and methoxy species on the Ni2+ sites. Additionally, the weak peak features at 1355, 1365 and 1578 cm−1 were detected. The peaks at 1355 and 1365 cm−1 are assigned to the symmetric O-C-O stretches, νs(OCO), of formate species, while the peak at 1578 cm−1 corresponds to the asymmetric O-C-O stretches, νas(OCO), of tilted formate species as reported in earlier studies [46]. Upon annealing to 400 K, the intensities of type-II and type-III methoxy peaks diminished, shifting to 1045 and 1020 cm−1, respectively, while the peaks for type-I methoxy species on the ceria surface and methoxy species on Ni2+ sites disappeared. Combined with the results of SRPES and TPD, the CO and H2 desorption at 300−430 K can be attributed to the decomposition of methoxy species adsorbed on Ni2+ sites. Concurrently, the intensities of νs(OCO) and νas(OCO) peaks at 1363 and 1584 cm−1 significantly increased. At 500 K, the νs(OCO) and νas(OCO) peaks for formate species vanished, leaving only the ν(CO) peaks for type-II and type-III methoxy species on the ceria surface. This observation supports the hypothesis that methoxy adsorbed on the bare ceria surface migrates to Ni2+ sites and decomposes to CO and H2 at 500–600 K. Consistent with results from the corresponding C 1s spectra (FIG. 2(b)), no formate species were detected due to the same annealing processes used in the XPS and IRRAS experiments. At 600 K, all carbon species were no longer detectable, consistent with previous SRPES results.

    Methanol decomposition on the Ni/CeO2(111) surface exhibits unique reaction pathways that are absent in both ceria and Ni single crystals. On the fully oxidized CeO2(111) surface, formaldehyde is the primary product of methanol decomposition, desorbing at temperatures between 500 and 600 K. In contrast, on a reduced ceria surface, formaldehyde production persists, but methanol predominantly decomposes into CO and H2 at ~650 K [23, 30]. On the pure Ni(111) surface, methanol decomposition similarly yields CO and H2 [42, 4752]. Notably, CO does not desorb immediately after its formation but requires annealing to 400–500 K. Conversely, H2 desorbs at ~220 K, a process influenced by the kinetic of hydrogen atoms recombination. Furthermore, the presence of surface oxygen increases the H2 desorption temperature to ~300 K and lowers the CO desorption temperature to ~350 K on the O-Ni(111) surface [42].

    Comparing methanol decomposition on ceria and Ni single crystal surfaces reveals new behavior on the Ni/CeO2(111) surface. The methanol adsorption at 300 K and its decomposition processes at different temperatures on the Ni/CeO2(111) surface are schematically shown in FIG. 5. Specifically, the oxygen vacancies form through the interaction between Ni and CeO2 on the surface, facilitating the methanol molecules dissociation into methoxy and hydroxyl groups, with a small portion of the methoxy converting into formate species. Our results indicate that Ni2+ species act as active sites for the low-temperature methoxy decomposition into CO and H2 between 300 and 430 K, with minor formaldehyde formation on edge sites of ceria.

    Figure  5.  The schematic diagram of methanol adsorption at 300 K and its decomposition processes at different temperatures on Ni/CeO2(111) surface.

    At Ni sites:

    CH3O-(a)CO(g)+3H(a)
    (1)
    2H(a)H2(g)
    (2)

    At edges sites of ceria:

    CH3O-(a)CH2O(g)+H(a)
    (3)

    At 500−600 K, some methoxy on the bare ceria surface migrate to the Ce-O-Ni mixed oxide, where they decompose into CO and H2. Additionally, the decomposition of methoxy on ceria terrace sites and the reaction of formaldehyde with hydrogen can produce formaldehyde within this temperature range.

    At Ni sites:

    CH3O-(a)CO(g)+3H(a)
    (4)
    2H(a)H2(g)
    (5)

    At terrace sites of ceria:

    CH3O-(a)CH2O(g)+H(a)
    (6)
    HCOO-(a)+H(a)CH2O(g)+O(lattice)
    (7)

    Understanding the catalyst deactivation and reactivation mechanism is crucial, particularly for enhancing the low-temperature activity of the Ni/CeO2(111) surface in methanol decomposition into CO and H2, compared to the pristine CeO2(111) surface. To investigate it, the C 1s and O 1s spectra of the 0.4 ML Ni/CeO2(111) surface were collected after stepwise annealing to 600 K with methanol for 30 min (FIG. 6(a) and FIG. S6 in SM). First, two peaks were evident at 286.7 eV (531.8 eV) and 289.2 eV (532.6 eV) for methoxy and formate species at 300 K, respectively [23, 24]. This suggests the methanol molecules dissociate into methoxy species and a portion of the methoxy species is converted into formate species. When the 0.4 ML Ni/CeO2(111) surface was exposed to methanol at 400 K, the peak for methoxy species shifted to 286.9 eV. Annealing at 500 K led to emergence of two peaks at 284.5 and 283.4 eV in the C 1s spectra, which are assigned to coke and Ni3C species [10]. At 600 K, only a distinct narrow peak at 284.5 eV assigned to surface coke was observed.

    Figure  6.  (a) C 1s spectra, =400 eV, and (b) Ni 2p spectra, =1486.6 eV, were recorded after annealing the 0.4 ML Ni/CeO2(111) surface at as indicated temperatures with 5×10−7 mbar methanol for 30 min. (c) The evolution of n(Ce3+)/(n(Ce3+)+n(Ce4+)) ratio, as determined from RPES, was recorded followed by annealing CeO2(111) (black line) and 0.4 ML Ni/CeO2(111) (red line) surfaces at as indicated temperatures with 5×10−7 mbar methanol.

    The corresponding Ni 2p spectra are shown in FIG. 6(b). Initially, the Ni 2p spectra remained stable after exposing it with methanol at 300 K. However, at 400 K an increasing intensity of peak at 852.7 eV and the slightly decreasing intensity of peak at 854.8 eV were observed indicating the partial reduction of Ni2+ species to metallic Ni0 [19, 21]. Upon annealing to 500 K, the absence of the peak at 854.8 eV and the higher peak at 852.7 eV indicate the complete reduction of Ni2+ species. Notably, the coke formation occurred simultaneously, which was not observed on the CeO2(111) surface (FIG. 6(a) and FIG. S6 in SM). This suggests that both coke and Ni3C species result from the dissociation of CO on metallic Ni0 sites located on the 0.4 ML Ni/CeO2(111) surface [53, 54]. At 600 K, the intensity of the Ni 2p peak diminished. Give that the sample had previously been annealed to 700 K, the effect of nanoparticle agglomeration can be excluded. Moreover, significant surface coke was detected in the corresponding C 1s spectra (FIG. 6(a)). This suggests that the reduction in the Ni 2p peak intensity is likely due to the Ni nanoparticles covered by surface coke. Moreover, the gradually increase of n(Ce3+)/(n(Ce3+)+n(Ce4+)) ratio at 300–500 K was observed, indicated the reduction of ceria due to the methanol reaction on the surface (FIG. 6(c)). However, the n(Ce3+)/(n(Ce3+)+n(Ce4+)) ratio remained stable upon annealing to 600 K. In contrast, annealing the CeO2(111) thin films with methanol from 300 K to 600 K continuously increases the n(Ce3+)/(n(Ce3+)+ n(Ce4+)) ratio (FIG. 6(c)). This difference is likely due to the significant formation of surface coke on the 0.4 ML Ni/CeO2 surface compared to the CeO2 surface at 600 K (FIG. 6(a) and FIG. S6 in SM). The reduction of ceria and Ni2+ species is likely attributed to two factors. One is the recombination of OH groups to form H2O during methanol decomposition [24, 31]. This is supported by the observation that the H2O is the main product in the TPD experiment (FIG. 3(b)). Additionally, the previous study confirmed the reaction of surface C with lattice O can happen in the methanol decomposition process, leading to the production of CO and contributing to the reduction of ceria and Ni2+species [10].

    To further investigate the impact of reduction of ceria and Ni2+ species, together with the formation of coke and Ni3C species during methanol decomposition on the Ni/CeO2(111) surface on catalytic performance, we analyzed the evolution of formation rates of two key products, CO and D2, in the course of annealing 0.4 ML Ni/CeO2(111) surface with CD3OD (FIG. 7). Upon annealing the sample from 300 K to 400 K, a noticeable increase in the formation rates of CO and D2 was observed, followed by a decline back to baseline levels observed before annealing. This behavior is attributed to the reduction of the Ni2+ species, underscoring the crucial role of Ni2+ species as the active site for low-temperature methoxy decomposition into CO and H2. Subsequently, to explore the influence of reverse oxygen spillover on the catalytic activity for methanol decomposition into CO and H2, the model catalyst underwent the annealing to 500 K. Noted that the formation rates of CO and D2 increased with temperature. This is likely due to the increase of Ni2+ species when the reactive temperature increases, which facilitates the oxygen reverse spillover from ceria to Ni. These rates plateaued at 500 K before gradually returning to the levels seen prior to annealing. Furthermore, the TPD experiment conducted on the as-prepared surface, which was annealed at 500 K with CD3OD, revealed the absence of the desorption peaks of CO and D2 at low temperatures (FIG. S7 in SM). There results suggest that annealing the Ni/CeO2(111) surface with methanol at 500 K reduces the catalytic activity for methanol decomposition into CO and H2, potentially leading to catalyst deactivation. By aligning these findings, we ascribe the deactivation process to the reduction of Ni2+ species to metallic Ni0 and the formation of coke and Ni3C species. Additionally, when 0.4 ML Ni/CeO2(111) surface is annealed with methanol at 500 K, the emergence of a plateau in the formation rates of CO and D2 indicates that the ceria support serves as an oxygen “reservoir”. This role can effectively delay the reduction of Ni2+ species and the formation of coke and Ni3C species via the reverse oxygen spillover mechanism [55, 56]. However, prolonged annealing at 500 K leads to a further reduction of ceria, thereby diminishing its oxygen buffering capacity.

    Figure  7.  Product formation rates of hydrogen (m/z=4), CO (m/z=28) from deuterated methanol decomposition on the 0.4 ML Ni/CeO2(111) surface at annealed as-indicated temperatures in a deuterated methanol background pressure of 1×10−8 mbar.

    The reduction of Ni2+ species and the formation of Ni3C species present significant challenge in the low-temperature methanol decomposition into CO and H2. To address this issue, we investigated the potential of O2 as a low-temperature activation agent for the reaction. Utilizing a 0.4 ML Ni/CeO2 model catalyst, we conducted an annealing process with methanol at 500 K for 30 min, followed by an annealing in O2 at the same temperature for an additional 30 min (FIG. S8 in SM). Notably, the C 1s spectra showed the absence of peaks assigned to methoxy and Ni3C species and the Ni 2p spectra revealed only a peak corresponding to Ni2+ species after post-treatment with O2, accompanied by a decrease in the ratio of n(Ce3+)/(n(Ce3+)+n(Ce4+)) to 0.7%. This indicates the re-oxidation of metallic Ni0, Ni3C species and ceria. It verifies the rejuvenation of the active sites, showcasing the potential of O2 for activation in low-temperature methanol decomposition into CO and H2.

    To validate the potential of O2 in activating sites on the Ni/CeO2(111) surface for low-temperature methanol decomposition into CO and H2, we employed QMS to measure the formation rates of D2 and CO after introducing O2 to the deactivated sample surface which had been annealed with CD3OD at 500 K (FIG. 8(a)). The introduction of O2 resulted in increase in the formation rates of D2 and CO. Notably, these product rates exhibited stable formation rates throughout the annealing process. Moreover, the results from IRRAS under similar reaction conditions are shown in FIG. 8(b). When only methanol was present, an IR peak at 1061 cm−1 was identified, corresponding to type-II methoxy species [25, 29]. However, the peak became imperceptible upon the addition of O2. Taken together, these experimental results suggest that the addition of O2 can reactivate sites for low-temperature methoxy decomposition, thereby facilitating the production of CO and H2.

    Figure  8.  (a) Product formation rates of hydrogen (m/z=4) and CO (m/z=28), examined following exposure the 0.4 ML Ni/CeO2(111) surface with 5×10−9 mbar O2 after annealing it with 1×10−8 mbar deuterated methanol at 500 K for 30 min. (b) IRRA spectra of as-prepared 0.4 ML Ni/CeO2(111) surface under 5×10−7 mbar methanol pressure (black line) and under oxidation conditions (red line) at 500 K.

    This study elucidates the complex dynamics of methanol decomposition on Ni/CeO2(111) surfaces, revealing distinct reaction pathways compared to those observed on the pristine CeO2(111) surface. Annealing the Ni/CeO2(111) surface to 700 K, leads to the formation of Ce-O-Ni mixed oxide as dominant phase due to the metal-support interaction. The Ni2+ species can promote the complete dehydrogenation of methoxy, originating from methanol dissociation on the Ni/CeO2 surface. At low temperatures (300–430 K), the majority of methoxy decomposes into CO and H2 on Ni2+ sites. As temperature increasing to 500–600 K, the methoxy species on the bare ceria surface migrate to the Ni2+ sites and subsequently decompose to produce CO and H2. However, annealing the catalyst under a methanol atmosphere at 500 K results in the reduction of Ni2+ species accompanied by the formation of coke and Ni3C species, which ultimately deactivates the Ni/CeO2 model catalyst. The study also demonstrates that the introduction of O2 can effectively reoxidize metallic Ni0 and Ni3C back to Ni2+ species and reactivate the deactivated catalyst. Overall, the findings highlight the critical importance of metal-support interactions as well as the role of reverse oxygen spillover in facilitating low-temperature methanol decomposition to CO and H2.

    Supplementary materials: The Ni 2p3/2 peak fitting of 0.4 ML Ni/CeO2 and 0.6 ML Ni/CeO2 surfaces which are annealed at 700 K; Ni 2p3/2 spectra of 0.4 ML NiO/CeO2(111); O 1s spectra and temperature-dependent normalized C 1s peak intensity analysis of methanol adsorbed on the CeO2(111) and 0.4 ML Ni/CeO2(111) surfaces and annealed at as indicated temperatures; the corrected signal of formaldehyde in the TPD spectra; C 1s and O 1s spectra obtained after annealing CeO2(111) surface at indicated temperatures with methanol; O 1s spectra obtained after annealing 0.4 ML Ni/CeO2(111) surface at as indicated temperatures with methanol; TPD spectra for 0.03 L of CD3OD at 105 K on the 0.4 ML Ni/CeO2(111) surface annealed with CD3OD at 500 K for 30 min; C 1s, Ni 2p, O 1s spectra and the n(Ce3+)/(n(Ce3+)+n(Ce4+)) ratio of 0.4 ML Ni/CeO2(111) surface were obtained by annealing it with methanol at 500 K for 30 min, following by annealing it with O2 at 500 K for 30 min are shown.

    This work was financially supported by the National Key R&D Program of China (2023YFA1509103) and the National Natural Science Foundation of China (Nos. 22272157, 21872131, 22106085, and U1932214).

  • [1]
    N. F. Dummer, D. J. Willock, Q. He, M. J. Howard, R. J. Lewis, G. D. Qi, S. H. Taylor, J. Xu, D. Bethell, C. J. Kiely, and G. J. Hutchings, Chem. Rev. 123, 6359 (2023). doi: 10.1021/acs.chemrev.2c00439
    [2]
    A. Kumar, P. Daw, and D. Milstein, Chem. Rev. 122, 385 (2022). doi: 10.1021/acs.chemrev.1c00412
    [3]
    M. H. Ren, Y. M. Zhang, X. Wang, and H. S. Qiu, Catalysts 12, 403 (2022). doi: 10.3390/catal12040403
    [4]
    X. Jiang, X. W. Nie, X. W. Guo, C. S. Song, and J. G. Chen, Chem. Rev. 120, 7984 (2020). doi: 10.1021/acs.chemrev.9b00723
    [5]
    A. Yuda, A. Ashok, and A. Kumar, Catal. Rev. 64, 126 (2022). doi: 10.1080/01614940.2020.1802811
    [6]
    S. Sá, H. Silva, L. Brandão, J. M. Sousa, and A. Mendes, Appl. Catal. B: Environ. 99, 43 (2010). doi: 10.1016/j.apcatb.2010.06.015
    [7]
    M. M. J. Li and S. C. E. Tsang, Catal. Sci. Technol. 8, 3450 (2018). doi: 10.1039/C8CY00304A
    [8]
    P. K. Panigrahi, B. Chandu, M. R. Motapothula, and N. Puvvada, Energy Fuels 38, 2630 (2024). doi: 10.1021/acs.energyfuels.3c04084
    [9]
    V. T. T. Phan, Q. P. Nguyen, B. Wang, and I. J. Burgess, J. Am. Chem. Soc. 146, 4830 (2024). doi: 10.1021/jacs.3c13222
    [10]
    Z. Y. Liu, S. Y. Yao, A. Johnston-Peck, W. Q. Xu, J. A. Rodriguez, and S. D. Senanayake, Catal. Today 311, 74 (2018). doi: 10.1016/j.cattod.2017.08.041
    [11]
    R. Pérez-Hernández, G. Mondragón-Galicia, A. Allende Maravilla, and J. Palacios, Phys. Chem. Chem. Phys. 15, 12702 (2013). doi: 10.1039/c3cp52032c
    [12]
    P. G. Lustemberg, R. M. Palomino, R. A. Gutiérrez, D. C. Grinter, M. Vorokhta, Z. Y. Liu, P. J. Ramírez, V. Matolín, M. V. Ganduglia-Pirovano, S. D. Senanayake, and J. A. Rodriguez, J. Am. Chem. Soc. 140, 7681 (2018). doi: 10.1021/jacs.8b03809
    [13]
    Y. Xie, J. J. Chen, X. Wu, J. J. Wen, R. Zhao, Z. L. Li, G. C. Tian, Q. L. Zhang, P. Ning, and J. M. Hao, ACS Catal. 12, 10587 (2022). doi: 10.1021/acscatal.2c02535
    [14]
    Q. Q. Tan, Z. S. Shi, and D. F. Wu, Int. J. Energy Res. 43, 5392 (2019). doi: 10.1002/er.4636
    [15]
    J. Chen, W. Li, and R. C. Shen, Korean J. Chem. Eng. 33, 500 (2016). doi: 10.1007/s11814-015-0189-5
    [16]
    C. X. Wang, L. Y. Zhang, W. W. Wang, F. Yang, and C. J. Jia, ACS Catal. 14, 9421 (2024). doi: 10.1021/acscatal.4c01561
    [17]
    A. Salcedo, P. G. Lustemberg, N. Rui, R. M. Palomino, Z. Y. Liu, S. Nemsak, S. D. Senanayake, J. A. Rodriguez, M. V. Ganduglia-Pirovano, and B. Irigoyen, ACS Catal. 11, 8327 (2021). doi: 10.1021/acscatal.1c01604
    [18]
    B. Q. Zhong, Z. W. Liang, J. Z. Huang, Y. C. Ye, Y. J. Liu, and T. J. Wang, Ind. Eng. Chem. Res. 63, 6521 (2024). doi: 10.1021/acs.iecr.3c03578
    [19]
    D. D. Kong, J. F. Zhu, and K. H. Ernst, J. Phys. Chem. C 120, 5980 (2016). doi: 10.1021/acs.jpcc.5b10338
    [20]
    S. D. Senanayake, J. Evans, S. Agnoli, L. Barrio, T. L. Chen, J. Hrbek, and J. A. Rodriguez, Top. Catal. 54, 34 (2011). doi: 10.1007/s11244-011-9645-6
    [21]
    Y. H. Zhou and J. Zhou, J. Phys. Chem. C 116, 9544 (2012). doi: 10.1021/jp300259y
    [22]
    E. W. Peterson, H. R. Wang, Z. Y. Liu, T. Ara, S. D. Senanayake, J. A. Rodriguez, and J. Zhou, J. Vac. Sci. Technol. A 40, 013209 (2022). doi: 10.1116/6.0001409
    [23]
    D. R. Mullins, M. D. Robbins, and J. Zhou, Surf. Sci. 600, 1547 (2006). doi: 10.1016/j.susc.2006.02.011
    [24]
    V. Matolín, J. Libra, M. Škoda, N. Tsud, K. C. Prince, and T. Skála, Surf. Sci. 603, 1087 (2009). doi: 10.1016/j.susc.2009.02.010
    [25]
    A. Siokou and R. M. Nix, J. Phys. Chem. B 103, 6984 (1999). doi: 10.1021/jp991127h
    [26]
    P. M. Albrecht and D. R. Mullins, Langmuir 29, 4559 (2013). doi: 10.1021/la400295f
    [27]
    R. M. Ferrizz, G. S. Wong, T. Egami, and J. M. Vohs, Langmuir 17, 2464 (2001). doi: 10.1021/la001729o
    [28]
    D. R. Mullins, S. D. Senanayake, and T. L. Chen, J. Phys. Chem. C 114, 17112 (2010). doi: 10.1021/jp103905e
    [29]
    J. J. Uhlrich, B. Yang, and S. Shaikhutdinov, Top. Catal. 57, 1229 (2014). doi: 10.1007/s11244-014-0296-2
    [30]
    J. Zhou and D. R. Mullins, J. Phys. Chem. B 110, 15994 (2006). doi: 10.1021/jp061985v
    [31]
    V. Matolín, V. Johánek, M. Škoda, N. Tsud, K. C. Prince, T. Skála, and I. Matolínová, Langmuir 26, 13333 (2010). doi: 10.1021/la101936b
    [32]
    K. Mašek, M. Škoda, J. Beran, M. Cabala, K. C. Prince, T. Skála, N. Tsud, and V. Matolín, Catal. Lett. 145, 1474 (2015). doi: 10.1007/s10562-015-1557-9
    [33]
    G. H. Li, S. W. Hu, Q. Xu, and J. F. Zhu, J. Phys. Chem. C 123, 23563 (2019). doi: 10.1021/acs.jpcc.9b07146
    [34]
    X. Cao, D. L. Zhang, X. W. Cheng, Q. Xu, L. J. Zhang, L. C. Huang, Y. Tu, X. Yu, T. F. Zhang, Y. W. Li, H. Niemantsverdriet, and J. F. Zhu, J. Phys. Chem. C 126, 21638 (2022). doi: 10.1021/acs.jpcc.2c07515
    [35]
    Y. Tu, L. C. Huang, X. W. Cheng, B. C. Tian, D. L. Zhang, J. Hu, H. H. Ding, Q. Xu, Y. F. Ye, and J. F. Zhu, J. Phys. Chem. Lett. 15, 4096 (2024). doi: 10.1021/acs.jpclett.4c00810
    [36]
    Q. Xu, X. W. Cheng, N. Q. Zhang, Y. Tu, L. H. Wu, H. B. Pan, J. Hu, H. H. Ding, J. F. Zhu, and Y. D. Li, Nano Res. 16, 8882 (2023). doi: 10.1007/s12274-023-5585-2
    [37]
    S. W. Hu, Y. Wang, W. J. Wang, Y. Han, Q. T. Fan, X. F. Feng, Q. Xu, and J. F. Zhu, J. Phys. Chem. C 119, 3579 (2015). doi: 10.1021/jp511691p
    [38]
    A. Neitzel, A. Figueroba, Y. Lykhach, T. Skála, M. Vorokhta, N. Tsud, S. Mehl, K. Ševčíková, K. C. Prince, K. M. Neyman, V. Matolín, and J. Libuda, J. Phys. Chem. C 120, 9852 (2016). doi: 10.1021/acs.jpcc.6b02264
    [39]
    L. Barrio, A. Kubacka, G. Zhou, M. Estrella, A. Martínez-Arias, J. C. Hanson, M. Fernández-García, and J. A. Rodriguez, J. Phys. Chem. C 114, 12689 (2010). doi: 10.1021/jp103958u
    [40]
    S. D. Senanayake and D. R. Mullins, J. Phys. Chem. C 112, 9744 (2008). doi: 10.1021/jp8016425
    [41]
    Y. Lykhach, M. Happel, V. Johánek, T. Skála, F. Kollhoff, N. Tsud, F. Dvořák, K. C. Prince, V. Matolín, and J. Libuda, J. Phys. Chem. C 117, 12483 (2013). doi: 10.1021/jp311008v
    [42]
    H. Öström, B. J. Zhang, T. Vallejo, B. Merrill, J. Huang, and J. LaRue, J. Chem. Phys. 156, 024704 (2022). doi: 10.1063/5.0072396
    [43]
    K. Mudiyanselage, A. E. Baber, Z. Y. Liu, S. D. Senanayake, and D. J. Stacchiola, Catal. Today 240, 190 (2015). doi: 10.1016/j.cattod.2014.06.001
    [44]
    W. O. Gordon, Y. Xu, D. R. Mullins, and S. H. Overbury, Phys. Chem. Chem. Phys. 11, 11171 (2009). doi: 10.1039/b913310k
    [45]
    H. E. Sanders, P. Gardner, and D. A. King, Surf. Sci. 331-333, 1496 (1995). doi: 10.1016/0039-6028(95)00284-7
    [46]
    J. Kubota, A. Bandara, A. Wada, K. Domen, and C. Hirose, Surf. Sci. 368, 361 (1996). doi: 10.1016/S0039-6028(96)01076-X
    [47]
    S. R. Bare, J. A. Stroscio, and W. Ho, Surf. Sci. 155, L281 (1985). doi: 10.1016/0039-6028(85)90003-2
    [48]
    J. N. Russell, I. Chorkendorff, and J. T. Yates, Surf. Sci. 183, 316 (1987). doi: 10.1016/S0039-6028(87)80213-3
    [49]
    S. J. Carey, W. Zhao, E. Harman, A. K. Baumann, Z. T. Mao, W. Zhang, and C. T. Campbell, ACS Catal. 8, 10089 (2018). doi: 10.1021/acscatal.8b02992
    [50]
    S. J. Carey, W. Zhao, A. Frehner, C. T. Campbell, and B. Jackson, ACS Catal. 7, 1286 (2017). doi: 10.1021/acscatal.6b02457
    [51]
    M. C. Wu, C. M. Truong, and D. W. Goodman, J. Phys. Chem. 97, 9425 (1993). doi: 10.1021/j100139a029
    [52]
    H. E. Sanders, P. Gardner, and D. A. King, Chem. Phys. Lett. 231, 481 (1994). doi: 10.1016/0009-2614(94)01270-9
    [53]
    A. V. Miller, V. V. Kaichev, I. P. Prosvirin, and V. I. Bukhtiyarov, J. Phys. Chem. C 117, 8189 (2013). doi: 10.1021/jp3122177
    [54]
    Z. Y. Liu, T. Duchoň, H. R. Wang, E. W. Peterson, Y. H. Zhou, S. Luo, J. Zhou, V. Matolín, D. J. Stacchiola, J. A. Rodriguez, and S. D. Senanayake, J. Phys. Chem. C 119, 18248 (2015). doi: 10.1021/acs.jpcc.5b04310
    [55]
    S. Das, J. Ashok, Z. Bian, N. Dewangan, M. H. Wai, Y. Du, A. Borgna, K. Hidajat, and S. Kawi, Appl. Catal. B: Environ. 230, 220 (2018). doi: 10.1016/j.apcatb.2018.02.041
    [56]
    B. B. Yang, J. Q. Xu, T. Tang, L. S. Jiang, K. H. Wu, Q. Zhang, M. L. Xie, H. J. Hu, and F. Guo, Chem. Eng. J. 497, 154748 (2024). doi: 10.1016/j.cej.2024.154748
  • Other Related Supplements

Catalog

    Figures(8)

    Article Metrics

    Article views PDF downloads Cited by()
    Related

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return