English

• Hydrogen peroxide (H₂O₂), one of the 100 most important chemicals in the world, has been recognized as an efficient and green oxidant because of its numerous advantages such as mild reaction conditions, fast reaction speed, high selectivity and pollution-free. As a result, hydrogen peroxide is heavily used as a feedstock or additive (e.g., oxidant [1,2], bleaching agent [3], disinfectant [4]) in a wide range of manufacturing sectors, such as chemicals [5], papermaking, textile [6], medicine [7], electronics, and environmental protection [8].

  The anthraquinone oxidation (AO) has maintained its position as the most mature hydrogen peroxide synthesis technology since 1840s [9,10], and still accounts for more than 95% of the world’s H₂O₂ production today. Despite the high H₂O₂ yield per cycle, the AO process also suffers from the complicated process, side reactions that lead to a considerable consumption of alkyl anthraquinone, rapid deactivation of hydrogenation catalyst and high separation cost [11]. Furthermore, this process is only profitable on a large-scale production (>4 × 10⁴ t/a) [12,13]. Given the existing issues in AO process, the direct synthesis of hydrogen peroxide from H₂ and O₂ over heterogeneous catalysts offers an attractive and green alternative process, which was first introduced in 1914 [14]. However, the process has not been commercialized yet due to the strong explosive potential of the mixture of H₂ and O₂ at high pressures, as well as the low productivity caused by the side reactions, such as hydrogenation and degradation of H₂O₂ in the process.

  The high activity and selectivity of Pd-based catalysts for direct synthesis of H₂O₂ is well known and has received great attention for decades [15]. Recent studies have highlighted the enhanced H₂O₂ selectivity by alloying Pd with a secondary metal such as Au. The latter is often attributed to the geometrical effect, which gives emphasis on the advantage of Pd monomers surrounded by Au atoms towards H₂O₂ formation, while contiguous Pd ensembles are more favorable for H₂O₂ hydrogenation [16,17]. As H₂O₂ synthesis over Pd-based bimetallic catalysts is a surface-catalyzed and structure-sensitive reaction, modification of catalyst surface composition via surface pretreatment may alter the chemical environment of active sites and the product selectivity.

  For selective H₂ oxidation on Pd catalysts, there are discrepancies regarding the active phases since the copresence of metal Pd (Pd⁰) and Pd oxide (PdO_x) in the reaction system. For example, Gudarzi et al. reported that PdO(Ⅱ) species showed a higher selectivity towards H₂O₂ than Pd⁰ in both mono Pd [18] and bimetallic Pd-Au [19] catalysts. The lower H₂O₂ selectivity on Pd⁰ was attributed to the reabsorption and the subsequent degradation of H₂O₂ on Pd⁰ surface according to Choudhary [20-25]. Contrarily, some studies reported higher conversion and selectivity toward H₂O₂ on partially reduced PdO_x species [26-28]. Recently, N-heterocyclic carbenes (NHC) was adopted as electronic modifiers of Pd, and the optimized catalyst with an appropriate surface Pd⁰/Pd²⁺ ratio yielded twice as much H₂O₂ formation rate as the unmodified counterpart [29].

  In Pd catalyzed H₂O₂ synthesis reaction, the chemical composition (e.g., Pd⁰/Pd²⁺ molar ratio) or particle size of the Pd species can be finely tuned during the calcination of the precursor and the pretreatment prior to reaction [30,31]. Calcination under prolonged duration or elevated temperature was found to result in an increase in the fraction of Pd²⁺ to formation of a surface oxide layer [32], and hence a decreased H₂O₂ formation rate [33]. But on the other hand, the increasing calcination temperature will undoubtedly be accompanied by a growing average crystallite size [34,35] and an affected H₂O₂ selectivity. More recently, Hang et al. reported a Pd/C catalyst calcined at 523 K showing a promising H₂O₂ productivity of 4443 mmol h⁻¹ g_Pd⁻¹ with 94.5% selectivity [36]. The enhanced catalytic performance was attributed to the formation of small and highly oxidized Pd NPs (1.4–2.5 nm). Similarly, Pritchard observed that a mild heat treatment at 575 K was capable of forming an optimal Pd⁰/Pd²⁺ ratio without causing a significant growth of particle size [37]. Simon et al. successfully encapsulated small Pd-rich NPs in the amorphous SnO_x layer through an oxidation-reduction-oxidation (O-R-O) heat treatment cycle, which prevents the sequential hydrogenation and degradation of H₂O₂ [38].

  In our previous work, we have known that small Pd NPs (particle size of 1.4–2.5 nm) and exposure of active sites at Pd⁰/PdO interfaces are important parameters enhancing H₂O₂ selectivity over Pd catalysts [39,40]. We also demonstrated that alloying Pd with appropriate amount of Au effectively reduced the quantity of contiguous Pd ensembles, thus inhibiting the degradation of H₂O₂ via hydrogenation pathway. The highest H₂O₂ selectivity of 48.1% was achieved with an optimal Pd/Au ratio of 2.0 and total metal content of 3.0 wt% [41]. On the basis of our previous studies on direct synthesis of H₂O₂ over Pd-based catalysts (Table S1 in Supporting information), this study demonstrated that catalyst calcination and pretreatment protocol had great impacts on the morphology and chemical composition of Pd NPs, and ultimately on catalytic performance with regard to selectivity and productivity of H₂O₂.

  The four catalyst samples were prepared and tested for their capability to catalyze the direct synthesis of H₂O₂ under atmospheric pressure. A prior blank experiment was used to exclude the catalytic activity from the support. Detailed catalyst preparation, characterization and testing methods are available in Supporting information. Since the formation of H₂O₂ always occurs with its hydrogenation/decomposition reactions, the H₂O₂ productivity is reported as its net formation rate [42]. Table 1 shows that H₂O₂ selectivity of unreduced catalyst remains 71.8%, but increases abruptly to 87.7% after reduction under 573 K. A further increase of reduction temperature to 673 K leads to a slight decrease in H₂O₂ selectivity. It is also noticed that catalysts reduced at mild temperatures (473 K and 573 K) exhibit higher H₂ conversion than the unreduced catalyst. Higher temperature reduction tends to trigger the sintering of the metal NPs, which may cause the reduction in the number of active sites as well as the activity loss. This is verified by the abrupt decrease in H₂ conversion when the reduction temperature increases to 673 K. A similar trend was observed for the H₂O₂ productivity as well. The highest H₂O₂ productivity was obtained with Pd_2.0Au_1.0-473 (3116.4 mmol h⁻¹ g_Pd⁻¹), which also exhibited the highest H₂ conversion (17.7%). By evaluating the dispersion of metal atoms exposed to reactants based on a hemispheric model, we also calculated the turnover frequency (TOF, h⁻¹), which showed a similar trend with the productivities. The TOFs of the reduced catalysts were generally higher than that of the unreduced one, with a highest value of 1103.3 h⁻¹ occurring on the catalysts reduced at the lowest temperature (473 K). And then, it showed a slight decrease after the first additional 100 K, but when it came to another same magnitude of temperature increase, it responded with a rapid decline of more than 30%.

  Table 1

  

  Table 1.  erformance of Pd_2.0Au_1.0/TiO₂ catalysts with different heat treatment temperature.^a

  DownLoad: CSV

  The impacts of H₂ reduction pretreatment on decomposition/hydrogenation side reactions of H₂O₂ were also studied. Fig. S1 (Supporting information) shows that the H₂O₂ concentration barely changes within 60 min under pure N₂ flow, indicating no activity for H₂O₂ decomposition, which agrees with our previous studies [41]. When switching the gas feed from N₂ to 15H₂/85N₂, a significant drop in H₂O₂ concentration was observed for all catalysts. Higher reduction temperature results in a faster H₂O₂ hydrogenation rate, which matches the growth of the average particle size measured by TEM (Fig. 1) which will be discussed in the next section. The enhanced hydrogenation activity on large Pd NPs is due to the fact that the cleavage of O-O bond is favored on the extended Pd⁰ ensembles. To sum up, the hydrogenation side reaction is the primary cause affecting the selectivity and net productivity of H₂O₂ during direct H₂O₂ synthesis from H₂ and O₂.

  Figure 1

  

  Figure 1.  TEM images and Pd particle sizes of the fresh catalysts. (a) Pd_2.0Au_1.0-0, (b) Pd_2.0Au_1.0-473, (c) Pd_2.0Au_1.0-573, (d) Pd_2.0Au_1.0-673.

  DownLoad: Full-Size Img PowerPoint

  The discrepancies in catalytic performance seen for Pd-Au catalysts with the same nominal composition but different reduction procedures encourage us to examine the surface/bulk structure and chemical composition of the catalysts for better understanding of the structure-performance relationship.

  Fig. 1 shows the TEM images of the Pd-Au bimetallic catalysts, as well as the corresponding particle size distribution of the Pd particles. It can be seen that the TiO₂ particles exhibit the same morphology after reduction pretreatment under different temperatures with an average particle size of 25 nm, suggesting that reduction treatment does not lead to obvious agglomeration of TiO₂ support. Pd NPs with different dimensions, ranging from small clusters of several atoms (<1 nm) to large Pd-Au particles (>10 nm), were clearly visualized on the images. Reduction pretreatment results in an abrupt increase in average Pd particle size from 2.8 ± 0.7 nm to 3.7 ± 0.8 nm. Increasing reduction temperature from 473 K to 673 K causes a moderate growth on the average particle size from 3.7 ± 0.8 nm to 4.2 ± 0.7 nm, equivalent to a growth rate of ca. 0.2–0.3 nm per 100 K. Analysis of the TEM images and particle size distribution indicate that reduction treatment under mild temepratures causes the agglomeration of metal NPs to some extent.

  High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) was used to analyze the micromorphology and the elemental distribution of Au and Pd (Fig. 2). The distributions of both Pd and Au are almost overlapped with that of Ti and O, confirming the homogeneity of the supported catalyst in the form of Pd-Au bimetallic particles. In the same figure, it can also be seen that the spatial distribution of aggregated forms of Au is highly consistent with that of large Pd NPs with a size of 10–15 nm, suggesting the presence of highly-enriched large particles in addition to the other highly-dispersed ones. The energy-dispersive X-ray spectroscopy (EDS) elemental mapping images confirmed a uniform distribution of Pd and Au in the bimetallic particles and the tight contact between the two metals. The hybrid support was also identified by the elemental mapping of Ti and O, as shown in Figs. 2b and c, respectively. The XRD patterns (Fig. 3a) of the both unreduced and reduced Pd-Au bimetallic catalysts reveal both characteristics of rutile (25.3°, 37.8° and 48.0°) and anatase phases (27.4°, 36.1° and 54.3°, Degussa P25) with identical 2θ positions, manifesting that the crystal structure of the TiO₂ support is intact during reduction pretreatment [43]. Pd shows a weak signal belonging to Pd (111) at 2θ = 40.1° in Pd_2.0Au_1.0-673 [44], whereas the diffraction peaks of Au are absent, presumably owing to the low metal loading and high metal dispersion. Such cases are not rare in palladium-based catalysts prepared by similar impregnation method [31]. Nevertheless, all catalysts are consisted of mainly small metal particles, in good agreement with the TEM results.

  Figure 2

  

  Figure 2.  Microstructure analysis of Pd_2.0Au_1.0/TiO₂ catalysts. (a) STEM-HAADF grayscale image of Pd_2.0Au_1.0/TiO₂, (b-f) STEM-EDS maps of Pd_2.0Au_1.0/TiO₂. Elements shown in blue, yellow, green and red refer to Ti, O, Pd and Au, respectively.

  DownLoad: Full-Size Img PowerPoint

  Figure 3

  

  Figure 3.  (a) XRD patterns, (b) CO-DRIFTS, (c) fitting Pd 3d XPS spectra, and (d) Au 4f XPS spectra of Pd_2.0Au_1.0/TiO₂ catalysts with different heat treatment temperature: (A) Pd_2.0Au_1.0-0, (B) Pd_2.0Au_1.0-473, (C) Pd_2.0Au_1.0-573, (D) Pd_2.0Au_1.0-673.

  DownLoad: Full-Size Img PowerPoint

  The surface characteristics of Pd-Au bimetallic catalysts after reduction treatment was studied using CO-DRIFTS, and the observed spectra were displayed in Fig. 3b. CO absorption bands located ~1800–2100 cm⁻¹ were observed in all catalysts, which were attributed to CO adsorbed on metallic Pd atoms with different bonds (e.g., linear, bridged, and hollow). No CO adsorption bands on Au (ca. 2100 cm⁻¹) was observed [45]. The peak at 2084 cm⁻¹ (peak 1) associated with the linearly bonded CO (Pd-CO) on Pd atoms of low coordination, e.g., corner or edge sites [46] is visible for all reduced catalysts, and this peak splits into two small peaks located at 2104 and 2078 cm⁻¹ for the unreduced catalyst [47]. The three successive peaks at 1985, 1936, and 1870 cm⁻¹ (peak 2) were assigned to chemisorbed CO on bridged- and multi-folded hollow Pd⁰ sites [48].

  More intriguingly, the bands related to CO adsorbed on these Pd⁰ sites grow sharper upon increasing reduction temperature. The peak area ratio of the CO_linear bands vs. CO_bridged + CO_hollow bands was decreased regularly from 0.059 to 0.025 (Table S2 in Supporting information), indicating that higher reduction pretreatment results in the exposure of more flat sites with continuous Pd ensembles. These features also shifted slightly to lower wavenumbers with increasing reduction temperatures, and eventually ended at 1976, 1890 and 1859 cm⁻¹ in the case of Pd_2.0Au_1.0-673. These redshifts may be attributed to a stronger electronic back-donation from the d orbital of Pd metal to the 2π molecular orbitals of CO molecules, indicating a remarkable enhancement in the d orbital electron density from the increase in Pd⁰ [49,50]. Moreover, a small band around 1607 cm⁻¹ on Pd_2.0Au_1.0-0 corresponding to surface carbonate species was absent in the spectra of the reduced catalysts presumably because the carbonate species were removed after reduction treatment.

  In Pd-catalyzed direct H₂O₂ synthesis, Pd⁰ is considered as responsible for H₂ dissociation while PdO is suggested to activate O₂ associatively. A balanced ratio of Pd⁰/Pd²⁺ is necessary for high H₂O₂ yield. Herein, the chemical states of the supported Pd NPs were studied via XPS survey of Pd 3d spectra, and the results were shown in Fig. 3c and Table 2. The curve-fitting of Pd 3d spectra show Pd⁰ (335.2 and 340.5 eV) and Pd²⁺ (336.8 and 342.1 eV) components for all catalyst samples [51,52]. The molar ratio of Pd⁰/Pd²⁺ increases steadily along with the rising reduction temperature, from 0.1 in the unreduced catalyst to 1.6 in Pd_2.0Au_1.0-673. The increased surface molar ratio of Pd⁰/Pd²⁺ after H₂ reduction treatment provides more sites for H₂ dissociation, thus resulting in a faster hydrogenation rate, as observed in Pd_2.0Au_1.0-473 [53,54]. However, excessive Pd⁰ would lead to a decrease in the PdAu-PdAuO_x interface sites as well as an increase in metallic Pd sites, which is more feasible for the breakage of O-O bond in O₂, as evidenced by theoretical calculation (Fig. S2 in Supporting information). The rising reduction temperature also leads to an adverse aggregation of both surface Pd and Au atoms, resulting in the gradual destruction of the high dispersion, which may dominate the downward catalyst performance, manifested as an even lower H₂O₂ productivity than the untreated catalysts when the reduction temperature reaches 673 K.

  Table 2

  

  Table 2.  The quantified XPS data of the surface Pd atoms.

  DownLoad: CSV

  For XPS Au 4f (Fig. 3d), the Au 4f_5/2 and Au 4f_7/2 BE peaks appeared at 87.2 eV and 83.5 eV, respectively, lower than those of naturally occurring metallic Au (BE 87.7 and 84.0 eV [55]) by 0.5 eV. This can be partly ascribed to the quantum size effect, which implies that the nano-sized Au particles have more surface atoms than bulk Au, resulting in a shift of the d-band towards Fermi level [56], as well as the electron transfer from Au⁰ to both Pd⁰ [57] and TiO₂ [58]. Upon increasing reduction temperature, the Au 4f peaks are further negatively shifted. It implies that high temperature reduction results in an increase in the electron density of Au.

  To get more insights on the effects of surface chemical composition on the adsorption behaviors of the Pd catalysts, O₂- and H₂-TPD studies were carried out and the results were shown in Figs. 4a and b, respectively. Two desorption peaks at 365 K and 447 K were detected for all catalysts, mainly attributed to the adsorption of O₂ molecules on oxygen vacancies of surface TiO₂ and Pd particles [59], respectively. The strong O₂ desorption peaks observed in the high temperature region of 850–950 K can be attributed to the decomposition of surface palladium oxide [60]. Upon rising reduction temperature, the molecular O₂ peaks on Pd particles gradually shifted to lower temperature with increased peak intensities. On the other hand, the high temperature O₂ desorption peak broadens and even flatten out in the Pd_2.0Au_1.0-673 sample, suggesting an increasing ratio of Pd⁰/Pd²⁺, which is in line with the XPS results. More importantly, it is known that the activated O₂ molecules are important for the outcome of the reaction, which means excessively strong adsorption may lead to oxygen dissociation while weak adsorption may lead to oxygen desorption from the catalyst surface. The 573 K reduced one seems to have the optimal adsorption strength, thus turning out with the highest H₂O₂ selectivity.

  Figure 4

  

  Figure 4.  (a) O₂-, (b) H₂-TPD profiles obtained for different catalysts in an Ar flow with a ramping rate of 10 K/min. (A) Pd_2.0Au_1.0-0, (B) Pd_2.0Au_1.0-473, (C) Pd_2.0Au_1.0-573, (D) Pd_2.0Au_1.0-673.

  DownLoad: Full-Size Img PowerPoint

  For H₂-TPD, all catalysts showed similar desorption profiles with two major desorption peaks, denoted as H_α (350–500 K) and H_β (620–670 K). The H_α peaks corresponding to the chemisorbed H₂ species desorption on metallic Pd sites contained two overlapped peaks. The peak centered at ca. 378 K was allocated to H₂ derived from β-hydride decomposition [61,62]; the peak centered at ca. 431 K can be attributed to the desorption of H₂ on isolated Pd⁰ sites that have no interactions with Au [61,63]. The H_β peaks were probably due to the H₂ desorption from highly dispersed Pd atoms in Pd-Au particles. Increasing reduction temperature results in a decrease of H_α desorption amount. Such behavior implies a loss of metallic Pd sites which may be ascribed to the particle growth upon H₂ reduction under elevated temperatures.

  A hydrogenation reaction pathway study in H₂O₂ synthesis on Pd catalysts through isotopic experiments evidenced an associative adsorption pathway for O₂ molecules [53], and the formation of OOH intermediate from the dissociated H atoms and activated O₂ molecules was considered as the key step for H₂O₂ synthesis [64]. Reaction pathway that leads to the dissociation of O₂ molecules tends to end up with H₂O as the main product. But the O-O bond cleavage, which was considered to be adverse for H₂O₂ formation, was significantly suppressed on Pd monomers surrounded by less active Au atoms [41,65]. Our previous study has already unraveled a Pd‒PdO‒TiO₂ interface where the Pd⁰ atoms were most exposed on the upper layers of particles, and PdO species were present at the interface of Pd NPs and TiO₂ [40]. In this work, we adopted the modeled PuAu-PdAuOx interface as the active sites for H₂O₂ synthesis, where the upper metallic Pd species take responsibility for H₂ dissociated activation and the oxidized Pd species for O₂ non-dissociated activation. The reducible support TiO₂ was selected to better ensure the coexistence of metallic and oxidized metal components on the catalyst surface and the highly dispersed Au serves to stabilize the enrichment of metal Pd at the top of the particles. Single metallic or oxidized Pd species will lead to a dissatisfactory H₂O₂ productivity. Because the initial structure of the interface between PdAu and PdAuO_x is very much dictated by the activation conditions, so as the particle size, the final catalytic performance will be completely different. These insights on the critical role of PdAu‒PdAuO_x interface in H₂O₂ synthesis inspired us to further refine the geometric and chemical properties of the interface for optimal H₂O₂ selectivity and yield by reduction pretreatment.

  The reduction pretreatment on Pd-Au bimetallic catalysts obviously outperforms the unreduced one in terms of both H₂ conversion and H₂O₂ selectivity. The low H₂ conversion of the Pd_2.0Au_1.0-0 sample can be attributed to a lack of Pd⁰ sites for H₂ dissociation, and the reduction of PdO_x by H₂ leads to a rapid accumulation of H₂O at the initial stage of the reaction, which undoubtedly compromises the average H₂O₂ selectivity. However, increasing reduction temperature not only leads to more Pd⁰ for H₂ dissociation, but also to a growth in particle size caused by varying degrees of metal sintering, as observed in the TEM results. And the latter explains the dropping H₂ conversion from 17.7% to 8.1% with the increasing reduction temperature, along with the H₂O₂ productivity.

  As mentioned above, H₂O₂ synthesis was primarily occurred on the PdAu‒PdAuO_x interface (Fig. 5a), the structure of which was affected by the reduction conditions. The surface Pd⁰/Pd²⁺ ratio of the original Pd_2.0Au_1.0/TiO₂ increases from 0.1 to 0.3 (reduced at 473 K) and 0.9 (reduced at 573 K). The optimal Pd⁰/Pd²⁺ ratio in the Pd_2.0Au_1.0-573 demonstrates the highest H₂O₂ selectivity due to the downward shift of the PdAu and PdAuO_x interface demarcation line (Fig. 5b) to form the maximum active sites towards H₂O₂ synthesis. Nonetheless, a further increase in reduction temperature brings an opposite effect. We owe this to a combined effect: (Ⅰ) The decreased PdAu‒PdAuO_x interface in step with the continues increased Pd⁰/Pd²⁺ ratio, directly reduced the number of active sites; (Ⅱ) The high rate of parallel side reaction caused by the uncontrollable particle size growth, for O₂ prefers to be dissociated on continuous Pd⁰ sites; (Ⅲ) The excessive tandem side reactions, as the hydrogenation tests have shown that contiguous Pd ensembles are more active for H₂O₂ hydrogenation than small Pd clusters [66].

  Figure 5

  

  Figure 5.  (a) The proposed reaction mechanism of H₂O₂ synthesis on Pd_2.0Au_1.0/TiO₂ catalysts. (b) The possible change of structural model after reduction treatment.

  DownLoad: Full-Size Img PowerPoint

  Based on the discussion above, it can be summarized that the hydrogenation activity and H₂O₂ selectivity depends on the balance of the sites for hydrogenation and associative O₂ activation, and it is highly sensitive to the reduction conditions (e.g., temperature). It should be also noted that the reduction temperature should be finely tuned to avoid the activity loss caused by the sintering of metal NPs.

  In conclusion, we successfully demonstrated an engineering control on the chemical states and geometry of PdAu‒PdAuO_x interface in a bimetallic Pd-Au/TiO₂ catalyst through a facile reduction pretreatment method, and eventually identified 573 K as the optimum reduction temperature. The optimal Pd⁰/Pd²⁺ ratio at about 0.9 observed in Pd_2.0Au_1.0-573 showed the highest H₂O₂ selectivity of 87.7% compared with the unmodified one (71.80%), which is able to locate the PdAu‒PdAuO_x interface closest to the largest diameter on the catalyst surface,thereby increasing the number of the exposed number of active sites despite particle agglomeration. The insights on valence and morphology effect on catalytic activity provides a feasible approach to improve the direct synthesis of H₂O₂ via the modification of the Pd active surface through pre-activation treatment.

  Declaration of competing interest

  We declare that we do not have any commercial or associative interest that represents a conflict of interest associated with the work submitted.

  Acknowledgments

  This work was supported by the National Natural Science Foundation of China (Nos. 91934302 and 22178110), the Dean/Opening Project of Guangxi Key Laboratory of Petrochemical Resource Processing and Process Intensification Technology (No. 2020K001) and the Project of the Department of Science and Technology of Sichuan Province (No. 2023YFQ0086).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2023.108446.

  
  
• 1. [1]

     N.P. Jayaweera, A.E. Hall, A.A. Wilson, et al., Inorg. Chim. Acta 506 (2020) 119507. doi: 10.1016/j.ica.2020.119507

  2. [2]

     M. Hajjami, A. Ghorbani-Choghamarani, M.A. Zolfigol, F. Gholamian, Chin. Chem. Lett. 23 (2012) 1323–1326. doi: 10.1016/j.cclet.2012.10.006

  3. [3]

     R. Hage, A. Lienke, Angew. Chem. Int. Ed. 45 (2006) 206–222. doi: 10.1002/anie.200500525

  4. [4]

     A.R. Matin, S. Yousefzadeh, E. Ahmadi, et al., Food Chem. Toxicol. 116 (2018) 129–137. doi: 10.1016/j.fct.2018.04.002

  5. [5]

     H.Y. Guo, J.C. Li, Y.L. Shang, Chin. Chem. Lett. 20 (2009) 1408–1410. doi: 10.1016/j.cclet.2009.06.037

  6. [6]

     A. Asghar, A.A. Abdul Raman, W.M.A. Wan Daud, J. Clean. Prod. 87 (2015) 826–838. doi: 10.1016/j.jclepro.2014.09.010

  7. [7]

     T. Hadzic, N. Aykin-Burns, Y. Zhu, et al., Free Radic. Biol. Med. 48 (2010) 1024–1033. doi: 10.1016/j.freeradbiomed.2010.01.018

  8. [8]

     Z. Wei, H. Xu, Z. Lei, et al., Chin. Chem. Lett. 33 (2022) 920–925. doi: 10.1016/j.cclet.2021.07.006

  9. [9]

     E. Santacesaria, M.D. Serio, A. Russo, U. Leone, R. Velotti, Chem. Eng. Sci. 54 (1999) 2799–2806. doi: 10.1016/S0009-2509(98)00377-7

  10. [10]

      X. Chen, J. Catal. 220 (2003) 254–257. doi: 10.1016/j.jcat.2003.07.007

  11. [11]

      T. Ishihara, Y. Ohura, S. Yoshida, et al., Appl. Catal. A Gen. 291 (2005) 215–221. doi: 10.1016/j.apcata.2005.01.045

  12. [12]

      J.K. Edwards, G.J. Hutchings, Angew. Chem. Int. Ed. 47 (2008) 9192–9198. doi: 10.1002/anie.200802818

  13. [13]

      J.K. Edwards, S.J. Freakley, A.F. Carley, C.J. Kiely, G.J. Hutchings, Acc. Chem. Res. 47 (2014) 845–854. doi: 10.1021/ar400177c

  14. [14]

      H. Henkel, W. Weber, Patent, US 1108752, 1914.

  15. [15]

      V.R. Choudhary, C. Samanta, P. Jana, Ind. Eng. Chem. Res. 46 (2007) 3237–3242. doi: 10.1021/ie0608408

  16. [16]

      J. Gong, Chem. Rev. 112 (2012) 2987–3054. doi: 10.1021/cr200041p

  17. [17]

      F. Gao, D.W. Goodman, Chem. Soc. Rev. 41 (2012) 8009–8020. doi: 10.1039/c2cs35160a

  18. [18]

      D. Gudarzi, W. Ratchananusorn, I. Turunen, T. Salmi, M. Heinonen, Top. Catal. 56 (2013) 527–539. doi: 10.1007/s11244-013-0014-5

  19. [19]

      D. Gudarzi, W. Ratchananusorn, I. Turunen, M. Heinonen, T. Salmi, Catal. Today 248 (2015) 58–68. doi: 10.1016/j.cattod.2014.04.011

  20. [20]

      C. Samanta, V.R. Choudhary, Catal. Commun. 8 (2007) 73–79. doi: 10.1016/j.catcom.2006.05.027

  21. [21]

      V.R. Choudhary, C. Samanta, T.V. Choudhary, Catal. Commun. 8 (2007) 1310–1316. doi: 10.1016/j.catcom.2006.11.035

  22. [22]

      C. Samanta, V.R. Choudhary, Appl. Catal. A Gen. 326 (2007) 28–36. doi: 10.1016/j.apcata.2007.03.028

  23. [23]

      C. Samanta, V.R. Choudhary, Catal. Commun. 8 (2007) 2222–2228. doi: 10.1016/j.catcom.2007.05.007

  24. [24]

      C. Samanta, V.R. Choudhary, Chem. Eng. J. 136 (2008) 126–132. doi: 10.1016/j.cej.2007.03.016

  25. [25]

      C. Samanta, V.R. Choudhary, Appl. Catal. A Gen. 330 (2007) 23–32. doi: 10.1016/j.apcata.2007.06.030

  26. [26]

      V.V. Krishnan, A.G. Dokoutchaev, M.E. Thompson, J. Catal. 196 (2000) 366–374. doi: 10.1006/jcat.2000.3036

  27. [27]

      R. Burch, P.R. Ellis, Appl. Catal. B Environ. 42 (2003) 203–211. doi: 10.1016/S0926-3373(02)00232-1

  28. [28]

      Q. Liu, K.K. Gath, J.C. Bauer, R.E. Schaak, J.H. Lunsford, Catal. Lett. 132 (2009) 342–348. doi: 10.1007/s10562-009-0104-y

  29. [29]

      R.J. Lewis, M. Koy, M. Macino, et al., J. Am. Chem. Soc. 144 (2022) 15431–15436. doi: 10.1021/jacs.2c04828

  30. [30]

      A.A. Herzing, A.F. Carley, J.K. Edwards, G.J. Hutchings, C.J. Kiely, Chem. Mater. 20 (2008) 1492–1501. doi: 10.1021/cm702762d

  31. [31]

      Y. Yan, H. Li, Z. Lu, et al., Chin. Chem. Lett. 30 (2019) 1153–1156. doi: 10.1016/j.cclet.2019.03.030

  32. [32]

      S. Melada, R. Rioda, F. Menegazzo, F. Pinna, G. Strukul, J. Catal. 239 (2006) 422–430. doi: 10.1016/j.jcat.2006.02.014

  33. [33]

      J.K. Edwards, J. Pritchard, M. Piccinini, et al., J. Catal. 292 (2012) 227–238. doi: 10.1016/j.jcat.2012.05.018

  34. [34]

      A. Villa, K. Dumoleijn, C. Evangelisti, K. Moonen, L. Prati, RSC Adv. 8 (2018) 15202–15206. doi: 10.1039/c8ra01987h

  35. [35]

      D.R. Fertal, M. Monai, L. Proaño, et al., Catal. Today 382 (2021) 120–129. doi: 10.1016/j.cattod.2021.08.005

  36. [36]

      V.T.T. Hang, Y.M. Chung, Korean J. Chem. Eng. 37 (2020) 65–71. doi: 10.1007/s11814-019-0417-5

  37. [37]

      J. Pritchard, M. Piccinini, R. Tiruvalam, et al., Catal. Sci. Technol. 3 (2013) 308–317. doi: 10.1039/C2CY20234D

  38. [38]

      S.J. Freakley, Q. He, J.H. Harrhy, et al., Science 351 (2016) 965–968. doi: 10.1126/science.aad5705

  39. [39]

      P. Tian, L. Ouyang, X. Xu, et al., J. Catal. 349 (2017) 30–40. doi: 10.1016/j.jcat.2016.12.004

  40. [40]

      L. Ouyang, P.F. Tian, G.J. Da, et al., J. Catal. 321 (2015) 70–80. doi: 10.1016/j.jcat.2014.10.003

  41. [41]

      L. Ouyang, G.J. Da, P.F. Tian, et al., J. Catal. 311 (2014) 129–136. doi: 10.1016/j.jcat.2013.11.008

  42. [42]

      E. Ntainjua N, M. Piccinini, J.C. Pritchard, et al., ChemSusChem 2 (2009) 575–580. doi: 10.1002/cssc.200800257

  43. [43]

      Y. Zhang, Z.R. Tang, X. Fu, Y.J. Xu, ACS Nano 4 (2010) 7303–7314. doi: 10.1021/nn1024219

  44. [44]

      Q. Meng, X. Wang, M. Xiao, et al., Chin. Chem. Lett. 34 (2023) 107221. doi: 10.1016/j.cclet.2022.02.026

  45. [45]

      T. Ricciardulli, J.S. Adams, M. DeRidder, et al., J. Catal. 404 (2021) 661–678. doi: 10.1016/j.jcat.2021.10.001

  46. [46]

      J. Xu, L. Ouyang, W. Mao, et al., ACS Catal. 2 (2012) 261–269. doi: 10.1021/cs200694k

  47. [47]

      F. Gao, Y. Wang, D.W. Goodman, J. Phys. Chem. C 113 (2009) 14993–15000. doi: 10.1021/jp9053132

  48. [48]

      T. Lear, R. Marshall, J.A. Lopez-Sanchez, et al., J. Chem. Phys. 123 (2005) 174706. doi: 10.1063/1.2101487

  49. [49]

      P. Hu, D.A. King, M.H. Lee, M.C. Payne, Chem. Phys. Lett. 246 (1995) 73–78. doi: 10.1016/0009-2614(95)01070-P

  50. [50]

      C. Lin, H. Pan, Z. Yang, et al., Ind. Eng. Chem. Res. 59 (2020) 6424–6434. doi: 10.1021/acs.iecr.9b05953

  51. [51]

      E.H. Voogt, A.J.M. Mens, O.L.J. Gijzeman, J.W. Geus, Surf. Sci. 373 (1997) 210–220. doi: 10.1016/S0039-6028(96)01180-6

  52. [52]

      F.P. Leisenberger, G. Koller, M. Sock, et al., Surf. Sci. 445 (2000) 380–393. doi: 10.1016/S0039-6028(99)01084-5

  53. [53]

      J.H. Lunsford, J. Catal. 216 (2003) 455–460. doi: 10.1016/S0021-9517(02)00070-2

  54. [54]

      S. Chinta, J. Catal. 225 (2004) 249–255. doi: 10.1016/j.jcat.2004.04.014

  55. [55]

      H.G. Boyen, G. Kastle, F. Weigl, et al., Science 297 (2002) 1533–1536. doi: 10.1126/science.1076248

  56. [56]

      F.W. Chang, H.Y. Yu, L. Selva Roselin, H.C. Yang, Appl. Catal. A Gen. 290 (2005) 138–147. doi: 10.1016/j.apcata.2005.05.024

  57. [57]

      R. Liu, Y. Yu, K. Yoshida, et al., J. Catal. 269 (2010) 191–200. doi: 10.1016/j.jcat.2009.11.007

  58. [58]

      R. Meyer, C. Lemire, S.K. Shaikhutdinov, H.J. Freund, Gold Bull. 37 (2004) 72–124. doi: 10.1007/BF03215519

  59. [59]

      M.A. Henderson, W.S. Epling, C.L. Perkins, C.H.F. Peden, U. Diebold, J. Phys. Chem. B 103 (1999) 5328–5337.

  60. [60]

      J.J.A. Hinojosa, H.H. Kan, J.F. Weaver, J. Phys. Chem. C 112 (2008) 8324–8331. doi: 10.1021/jp800216x

  61. [61]

      V.H. Sandoval, C.E. Gigola, Appl. Catal. A Gen. 148 (1996) 81–96. doi: 10.1016/S0926-860X(96)00224-4

  62. [62]

      J. Lee, H. Rhee, J. Catal. 177 (1998) 208–216. doi: 10.1006/jcat.1998.2100

  63. [63]

      P.J. Berlowitz, D.W. Goodman, Langmuir 4 (1988) 1091–1095. doi: 10.1021/la00083a004

  64. [64]

      M.G. Seo, H.J. Kim, S.S. Han, K.Y. Lee, Catal. Surv. Asia 21 (2017) 1–12. doi: 10.1007/s10563-016-9221-y

  65. [65]

      H.C. Ham, G.S. Hwang, J. Han, S.W. Nam, T.H. Lim, J. Phys. Chem. C 113 (2009) 12943–12945. doi: 10.1021/jp9043545

  66. [66]

      V.R. Choudhary, C. Samanta, T.V. Choudhary, Appl. Catal. A Gen. 308 (2006) 128–133. doi: 10.1016/j.apcata.2006.04.010

• 

  Figure 1  TEM images and Pd particle sizes of the fresh catalysts. (a) Pd_2.0Au_1.0-0, (b) Pd_2.0Au_1.0-473, (c) Pd_2.0Au_1.0-573, (d) Pd_2.0Au_1.0-673.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Microstructure analysis of Pd_2.0Au_1.0/TiO₂ catalysts. (a) STEM-HAADF grayscale image of Pd_2.0Au_1.0/TiO₂, (b-f) STEM-EDS maps of Pd_2.0Au_1.0/TiO₂. Elements shown in blue, yellow, green and red refer to Ti, O, Pd and Au, respectively.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) XRD patterns, (b) CO-DRIFTS, (c) fitting Pd 3d XPS spectra, and (d) Au 4f XPS spectra of Pd_2.0Au_1.0/TiO₂ catalysts with different heat treatment temperature: (A) Pd_2.0Au_1.0-0, (B) Pd_2.0Au_1.0-473, (C) Pd_2.0Au_1.0-573, (D) Pd_2.0Au_1.0-673.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (a) O₂-, (b) H₂-TPD profiles obtained for different catalysts in an Ar flow with a ramping rate of 10 K/min. (A) Pd_2.0Au_1.0-0, (B) Pd_2.0Au_1.0-473, (C) Pd_2.0Au_1.0-573, (D) Pd_2.0Au_1.0-673.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  (a) The proposed reaction mechanism of H₂O₂ synthesis on Pd_2.0Au_1.0/TiO₂ catalysts. (b) The possible change of structural model after reduction treatment.

  下载: 全尺寸图片 幻灯片

  Table 1.  erformance of Pd_2.0Au_1.0/TiO₂ catalysts with different heat treatment temperature.^a

  下载: 导出CSV

  Table 2.  The quantified XPS data of the surface Pd atoms.

  下载: 导出CSV
•