PdPt y /V 2 O 5 -TiO 2 : Highly Active Catalysts with Good Moisture- and Sulfur Dioxide-Resistant Performance in Toluene Oxidation

: Catalytic performance and moisture and sulfur dioxide resistance are important for a catalyst used for the oxidation of volatile organic compounds (VOCs). Supported noble metals are active for VOC oxidation, but they are easily deactivated by water and sulfur dioxide. Hence, 2 , thus making the 0.46PdPt 2.10 /V 2 O 5 TiO 2 catalyst show good sulfur dioxide resistance.


Introduction
Volatile organic compounds (VOCs) are important precursors of PM 2.5 and ozone that lead to atmospheric environmental problems, such as haze and photochemical smog. In the emitted VOCs, however, there are some toxic gases such as SO 2 [1]. Hence, many countries have enacted strict laws to control VOC emissions [2,3]. Among all of the methods to eliminate VOCs, catalytic oxidation is considered to be an effective pathway owing to its high removal efficiency, energy saving, and no secondary pollution [4,5]. The key issue of such a technology is the development of an efficient and stable catalyst.
It was reported that the supported noble metal catalysts were conducive to the adsorption and activation of VOCs and oxygen, thus exhibiting high low-temperature activities  Figure 2 shows TEM and HAADF-STEM images and elemental mappings of the samples as well as EDX line scan analysis of the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample. The HADDF-STEM images and EDX elemental mappings of the as-obtained samples showed that the Pt, Pd, and PdPt y NPs were highly dispersed on the surface of V 2 O 5 -TiO 2 . We also obtained line scanning data of the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample, which showed that a bimetallic metal catalyst was successfully prepared. The SEM images of V 2 O 5 -TiO 2 are shown in Figure S1 (Supplementary Materials); the BET surface area, actual noble metal loadings, and Pt/Pd molar ratios of the samples are provided in Table 1; and the nitrogen adsorption-desorption isotherms and pore-size distributions of the samples are provided in Table 1 and Figure S2, respectively. It can be observed that surface areas (25.6-28.8 m 2 /g) of the supported noble metal samples were slightly decreased, as compared with that (33.0 m 2 /g) of V 2 O 5 -TiO 2 .

Catalytic Performance
By using a gas chromatograph (GC-2014C, Shimadzu) with a FID, we detected the products of toluene oxidation over the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample at different temperatures and SV = 40,000 mL/(g h), and their gas chromatogram curves are shown in Figure S3. Obviously, there was only one signal that was assignable to toluene, and no other products were detected. In other words, no by-products (in addition to CO 2 and H 2 O) were detected in the oxidation of toluene; i.e., CO 2 and H 2 O selectivities were almost 100%. It should be noted that the intermediates with very low concentrations produced during the toluene oxidation could be detected using the in situ DRIFTS technique, which is more sensitive than the GC technique in detecting very low amounts of organics.
Catalytic toluene oxidation activities and Arrhenius plots of the samples are shown in Figure 3. Obviously, toluene conversion over each sample increased with the rise in temperature. The catalytic performance of V 2 O 5 -TiO 2 after loading of Pt, Pd, or PdPt y NPs was remarkably enhanced, as compared with that of the V 2 O 5 -TiO 2 support. Especially, the supported PdPt y bimetallic samples showed excellent catalytic activities. For the sake of convenient comparison, T 50% and T 90% (the reaction temperatures required when toluene conversion reaches 50% and 90%, respectively) are used to evaluate the catalytic activity. As can be seen from Figure 3A, 0.46PdPt 2.10 /V 2 O 5 -TiO 2 performed the best (T 50% = 220 • C and T 90% = 245 • C at SV = 40,000 mL/(g h)).

Catalytic Stability and H 2 O, CO 2 , and SO 2 Resistance
There have been many reports on the effect of water vapor on catalytic activity [33]. For example, Zhang et al. [34] explored the effect of water on the methane oxidation activity of 0.44PtPd 2.20 /ZrO 2 and observed that the catalytic activity decreased owing to the introduction of water. In addition to the thermal stability, we also examined the effect of water vapor on the catalytic performance of 0.46PdPt 2.10 /V 2 O 5 -TiO 2 at SV = 40,000 mL/(g h), as shown in Figure 4A. Apparently, no significant changes in activity took place after 5.0 vol% water vapor was added to the reaction system. This demonstrates that the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample possesses good water resistance. That is to say, toluene was more strongly adsorbed on the catalyst surface than water vapor, resulting in the catalytic activity not being affected significantly. CO 2 is one of the products in toluene oxidation, and it can also influence the stability of a catalyst. For instance, Fu et al. [35] introduced 5.0 vol% CO 2 to the toluene combustion system over the 0.37Pt-0.16MnO x /meso-CeO 2 catalyst and observed that toluene conversion decreased slightly and such a partial deactivation was reversible. To explore the effect of CO 2 on catalytic activity, we introduced 5.0 vol% CO 2 to pass through the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample at 245 • C and SV = 40,000 mL/(g h) and measured the resulting activities, as shown in Figure 4B. Obviously, the catalytic activity of toluene oxidation over 0.46PdPt 2.10 /V 2 O 5 -TiO 2 was not decreased significantly, which might be due to the fact that the alkalinity of the TiO 2 support is weaker [36], and the adsorption of CO 2 on TiO 2 and/or PdPt 2.10 NPs is also weaker. That is to say, the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample exhibited good CO 2 resistance.
In some cases, there is the presence of a small amount of SO 2 in VOCs, which can poison the catalysts, thus resulting in an irreversible deactivation of the catalysts. After having monitored the changes in the activity of Pt/3DOM Mn 2 O 3 for toluene combustion after introducing 40 ppm sulfur dioxide to the reaction system, Pei et al. [37] found that toluene conversion decreased significantly, which was due to the formation of the sulfate species that occupied the active sites. In other words, SO 2 addition induced an irreversible deactivation of the catalyst. One of the main aims of the present work is to prepare a support with good SO 2 resistance so that the sulfur resistance of the catalyst could be improved by competition for the adsorption of SO 2 with the active noble metal sites. As shown in Figure 4C-E, toluene oxidation experiments were conducted over V 2 O 5 -TiO 2 , 0.47Pt/V 2 O 5 -TiO 2 , and 0.46PdPt 2.10 /V 2 O 5 -TiO 2 in the presence of 50 ppm SO 2 at SV = 40,000 mL/(g h).
Obviously, toluene conversions first dropped slightly and then increased slightly over the V 2 O 5 -TiO 2 support after the introduction of 50 ppm SO 2 . After 50 ppm SO 2 was introduced, toluene conversion over 0.47Pt/V 2 O 5 -TiO 2 decreased rapidly, which was associated with the fact that SO 2 is adsorbed at the active Pt site, thus resulting in a decrease in active site amount for the adsorption of toluene and/or O 2 . That is to say, SO 2 competed for the adsorption with reactants (toluene and oxygen). After about 4 h of reaction, the competitive adsorption of SO 2 and reactants at the active Pt site reached equilibrium, and toluene conversion decreased to 45%. When 50 ppm SO 2 was removed from the reaction system, toluene conversion was recovered to 90%. Over the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample, however, a toluene conversion of 90% was still maintained after 50 ppm SO 2 was introduced, and the activity was restored to its initial conversion level when SO 2 was cut off. This result indicates that the support plays an important role in the sulfur resistance of a catalyst. The catalytic activity of 0.46PdPt 2.10 /V 2 O 5 -TiO 2 decreased slightly within 8 h of on-stream toluene oxidation at 245 • C after the addition of 50 ppm SO 2 and could return to its initial activity level after SO 2 provision was cut off. Therefore, we conclude that the use of the acidic V 2 O 5 -TiO 2 support as the main site of sulfur dioxide species adsorption can effectively improve the sulfur dioxide resistance of 0.46PdPt 2.10 /V 2 O 5 -TiO 2 .  We also investigated the thermal stability and sulfur dioxide resistance of the 0.46PdPt 2.10 / V 2 O 5 -TiO 2 sample, and the results are shown in Figure 4F,G. Obviously, no significant drops in toluene conversion were observed within 25 h of on-stream reaction at 245 • C over 0.46PdPt 2.10 /V 2 O 5 -TiO 2 , indicating that the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample was stable under the dry and wet conditions. The TGA technique was used to analyze weight losses of the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 and 0.47Pt/V 2 O 5 -TiO 2 samples before and after 25 h of toluene oxidation in the presence of 50 ppm SO 2 at SV = 40,000 mL/(g h), and their TGA curves are shown in Figure S4. It can be seen that there was a gradual weight loss of ca. 0.7 wt% in the range of RT-600 • C for the fresh and used 0.46PdPt 2.10 /V 2 O 5 -TiO 2 samples, which was due to the removal of the adsorbed reactants and CO 2 and H 2 O; no weight loss due to the removal of the strongly adsorbed SO 2 on the sample surface was detected, demonstrating that the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample possessed a good resistance to SO 2 . For the 0.47Pt/V 2 O 5 -TiO 2 sample after the sulfur dioxide treatment, however, a weight loss of ca. 0.7 wt% was observed in the range of 300-470 • C, which was owing to the decomposition of the strongly adsorbed SO 2 or the formed sulfite species, indicating that the 0.47Pt/V 2 O 5 -TiO 2 sample tended to be poisoned by SO 2 .
To gain information on the physicochemical property change of the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample before and after 25 h of toluene oxidation in the presence of 50 ppm SO 2 at an SV of 40,000 mL/(g h), we measured the sample's XRD patterns, as shown in Figure  S5. It can be seen that the crystal phase structure of the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample was not altered before and after 25 h of toluene oxidation in the presence of 50 ppm SO 2 , which was in consistency with its stable catalytic performance. That is to say, the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample was durable under the adopted reaction conditions.

Surface Property
V 2p, Ti 2p, O 1s, S 2p, and Pt 4f XPS spectra of the fresh and SO 2 -treated 0.47Pt/V 2 O 5 -TiO 2 and 0.46PdPt 2.10 /V 2 O 5 -TiO 2 samples are shown in Figure 5, and their surface elemental compositions are listed in Table 3. The V 2p XPS spectrum of each sample was deconvoluted into three components at binding energy (BE) = 517.2, 515.9, and 515.4 eV ( Figure 5A), which belonged to the surface V 5+ , V 4+ , and V 3+ species [38][39][40], respectively. After SO 2 treatment, the V 5+ /V γ+ molar ratio decreased on 0.46PdPt 2.10 /V 2 O 5 -TiO 2 and 0.47Pt/V 2 O 5 -TiO 2 ; i.e., there were greater amounts of V 3+ and V 4+ species on both SO 2treated samples. As shown in Table 3, the S 6+ /S 4+ molar ratio (1.95) on the SO 2 -treated 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample was much higher than that (1.06) on the SO 2 -treated 0.47Pt/V 2 O 5 -TiO 2 sample, indicating that there was a greater amount of sulfate species on the former than on the latter. The results suggest that electrons can be transferred from V to S (i.e., V 5+ + S 4+ ⇔ V γ+ + S 6+ ), indicating that after SO 2 treatment, vanadium could adsorb SO 2 to form vanadium sulfite and/or sulfate, thus protecting the active noble metal site to a certain extent. The Ti 2p XPS spectrum of each sample was divided into two components at BE = 458.4 and 464.1 eV ( Figure 5B), which were assignable to the characteristic signal of the Ti 4+ species [41]. Each of the O 1s XPS spectra could be decomposed into three components at BE = 529.7, 531.3, and 533.0 eV ( Figure 5C), ascribable to the surface lattice oxygen (O latt ), adsorbed oxygen (O ads ), and adsorbed molecular water or carbonate species [42][43][44], respectively. As shown in Figure 5D, the two components at BE = 168.5 and 169.7 eV were attributable to the surface S 6+ and S 4+ species [45], respectively. After making the quantitative analysis, we can see that the 0.47Pt/V 2 O 5 -TiO 2 sample after SO 2 treatment shows a greater amount of sulfite and a lesser amount of sulfate, whereas the amounts of sulfite and sulfate on the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample after SO 2 treatment are just the opposite.  After curve fitting the Pt 4f spectrum of each sample, we can see three components at BE = 74.9, 75.8, and 78.2 eV ( Figure 5E), among which the one at BE = 75.8 eV was ascribed to the surface oxidized Pt (Pt 2+ ) species, while the ones at BE = 74.9 and 78.2 eV were classified as the surface oxidized Pt (Pt 4+ ) species [46][47][48][49]. According to the quantitatively analyzed data listed in Table 3, we can see that the O ads /O latt molar ratio on the SO 2treated 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample decreased only slightly, and the Pt 2+ /Pt 4+ molar ratio increased slightly but the Pd 0 /Pd 2+ molar ratio increased remarkably from 0.25 to 0.43; however, the O ads /O latt molar ratio on the SO 2 -treated 0.47Pt/V 2 O 5 -TiO 2 sample decreased considerably, and the Pt 2+ /Pt 4+ molar ratio increased markedly. The change in O ads /O latt , Pd 0 /Pd 2+ , or Pt 2+ /Pt 4+ molar ratio on the samples after SO 2 treatment might be associated with the adsorbed SO 2 species. Similarly, the Pd 3d spectrum of each Pd-containing sample was divided into four components: the two components at BE = 334.9 and 340.4 eV ( Figure 5F) were assigned to the surface metallic Pd (Pd 0 ) species, whereas the other two components at BE = 337.5 and 342.7 eV were attributed to the surface oxidized Pd (Pd 2+ ) species [50,51]. The Pd 0 /Pd 2+ molar ratio increased because the sulfur dioxide could react with the surface lattice oxygen of PdO to form sulfur trioxide, which further consumed the lattice oxygen on the PdO surface [52]. In other words, the reaction of SO 2 with the O latt on the PdO surface could promote the local reduction of PdO, resulting in an increase in the amount of the Pd 0 species on the sample surface.
The V 2p, Ti 2p, Pt 4f, Pd 3d, and O 1s XPS spectra of the 0.39Pd/V 2 O 5 -TiO 2 , 0.49PdPt 0.44 /V 2 O 5 -TiO 2 , and 0.41PdPt 0.85 /V 2 O 5 -TiO 2 samples are shown in Figure S6, and their surface elemental compositions are listed in Table 3. We notice that the O ads /O latt molar ratios on the PdPt y -loaded samples were higher than that on the Pd-or Pt-loaded sample, with the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample exhibiting the highest O ads /O latt molar ratio (Table 3). That is to say, the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample possessed the strongest oxygen activation ability. In a sulfur-containing atmosphere, vanadium can compete with precious metals to adsorb SO 2 and protect the active Pt site from being poisoned by SO 2 . It is hence understandable that the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample showed the highest activity in catalyzing the oxidation of toluene.

Reducibility and Oxygen Mobility
It is well known that the activity of a catalyst is associated with its low-temperature reducibility. Shown in Figure 6A are H 2 -TPR profiles of the V 2 O 5 -TiO 2 , 0.39Pd/V 2 O 5 -TiO 2 , 0.47Pt/V 2 O 5 -TiO 2 , 0.49PdPt 0.44 /V 2 O 5 -TiO 2 , 0.41PdPt 0.85 /V 2 O 5 -TiO 2 , and 0.46PdPt 2.10 /V 2 O 5 -TiO 2 samples. Two reduction peaks were detected at 529 and 650 • C for the V 2 O 5 -TiO 2 support: the former was attributed to the reduction of the surface V 5+ and Ti 4+ to the V 4+ and Ti 3+ species, while the latter was ascribable to the reduction of the bulk V 2 O 5 and TiO 2 species [53][54][55]. With the loading of Pt, Pd, or PdPt y NPs, the position of the reduction peak was shifted to a lower temperature, which was owing to the reduction of the oxidized noble metal in the supported sample. These results indicate that there is a strong interaction between Pt, Pd, or PdPt y NPs and V 2 O 5 -TiO 2 . According to the quantitative analysis data (Table 3), we can realize that the loading of Pt, Pd, or PdPt y NPs on V 2 O 5 -TiO 2 causes the H 2 consumption of the sample to increase significantly. The reduction temperature of the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample was the lowest, indicating that this sample possessed the best low-temperature reducibility. Oxygen desorption behaviors of the samples were measured using the O 2 -TPD technique, and their profiles are illustrated in Figure 6B and Figure S7. The peak in the low-temperature (<300 • C) region was considered as desorption of the chemically adsorbed O ads (O 2 − , O 2 2− , or O − ) species and decomposition of the PtO x , PdO x , and PtPdO x species; the one in the range of 300-600 • C was classified as desorption of the surface O latt species on V 2 O 5 -TiO 2 ; and the one above 600 • C was regarded as desorption of the bulk O latt species in V 2 O 5 -TiO 2 [56]. Catalytic oxidation of toluene usually occurs below 600 • C. Hence, the main active oxygen species are the O ads species and the O latt species in noble metal oxides that were desorbed at lower temperatures. Of course, it cannot be ruled out that the surface O latt species on V 2 O 5 -TiO 2 can participate in the oxidation of toluene at higher temperatures. Additionally, it should be noticed that the temperatures of oxygen desorption from the supported noble metal samples were shifted to lower temperatures as compared with those of the V 2 O 5 -TiO 2 support.

Toluene and Sulfur Dioxide Adsorption Behaviors
The adsorption of toluene on the catalyst is an important factor affecting its catalytic performance. The toluene-TPD experiments of the 0.47Pt/V 2 O 5 -TiO 2 and 0.46PdPt 2.10 /V 2 O 5 -TiO 2 samples were carried out, and their desorption profiles are shown in Figure 7A. For each sample, there was one toluene desorption peak with weak chemisorption at 90 • C. The 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample possessed the largest toluene desorption, suggesting that the sample can adsorb the largest amount of toluene, which is beneficial for the improvement in catalytic toluene oxidation activity. Figure 7B,C show MS spectra of carbon dioxide and water. It can be seen that when toluene is adsorbed on the sample surface, two different types of oxygen participate in the reaction. The peak at 90-300 • C was due to the reaction of the surface adsorbed oxygen species with toluene, while the one at 600-700 • C was owing to the reaction of the surface lattice oxygen species with toluene. The results were consistent with those of O 2 -TPD characterization. (C 7 H 8 + SO 2 )-TPD was used to investigate the thermal decomposition of the sulfate species on the sample surface after reaction, and their profiles are presented in Figure 7D. The V 2 O 5 -TiO 2 , 0.47Pt/V 2 O 5 -TiO 2 , and 0.46PdPt 2.10 /V 2 O 5 -TiO 2 samples exhibited three desorption peaks. The desorption peak at 386 • C was due to the decomposition of the V 2 (SO 4 ) 3 species, the one at 436 • C was assigned to the decomposition of the VOSO 4 species, and the one at 588 • C was ascribed to the decomposition of the TiOSO 4 species [57]. It is well known that the desorption peak below 400 • C is due to the decomposition of the weakly chemisorbed SO 2 and sulfite species, and the desorption peak above 400 • C is attributable to the decomposition of the strongly adsorbed sulfate species [58]. As revealed by the XPS characterization, the sulfate or sulfite species were formed on the surface of the sample after SO 2 treatment. The loading of precious metal NPs can make SO 2 be readily oxidized to SO 3 . Additionally, vanadium competes with precious metals to adsorb sulfur oxide species, which can protect the active noble metal site from being poisoned by SO 2 or SO 3 to a certain extent. This means that the sulfate species produced at the vanadium site are increased. From the (C 7 H 8 + SO 2 )-TPD results, we can see that the 0.47Pt/V 2 O 5 -TiO 2 sample possessed the largest amount of sulfur species desorption below 400 • C, which might be mainly due to the weak chemical adsorption of SO 2 on the Pt site, thus causing the sample to show a lower catalytic activity. Due to the enhanced Pt-Pd interaction, it was difficult for SO 2 to be adsorbed at the precious metal sites in the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample, thus protecting the active noble metal sites. In addition, we notice that the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample strongly adsorbs sulfur species to form vanadium sulfite or sulfate, and its desorption amount was larger as compared with that above 400 • C of the 0.47Pt/V 2 O 5 -TiO 2 sample, giving rise to a slight drop in catalytic activity. The desorption peak of the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 or 0.47Pt/V 2 O 5 -TiO 2 sample at about 370 • C significantly increased in intensity as compared with that of the V 2 O 5 -TiO 2 support. It is speculated that the catalyst may catalyze the oxidation of toluene in the sulfur-containing atmosphere, and vanadium is the main adsorption site of the sulfur species. Therefore, toluene conversion over the 0.47Pt/V 2 O 5 -TiO 2 sample was recovered to 90% after cutting off sulfur dioxide, while the activity of 0.46PdPt 2.10 /V 2 O 5 -TiO 2 remained basically unchanged in the sulfur-containing atmosphere.

Adsorption Mechanisms of Toluene and Sulfur Dioxide
In order to determine whether the sulfate species are produced on the surface of the sample, we collected FT-IR spectra of the 0.47Pt/V 2 O 5 -TiO 2 ( Figure 8A) and 0.46PdPt 2.10 /V 2 O 5 -TiO 2 ( Figure 8B) samples before and after toluene oxidation in the presence of 50 ppm SO 2 at SV = 40,000 mL/(g h) for 8 h. According to the assignments of absorption bands reported in the literature, we ascribe the band at 1115 cm −1 to the stretching vibration of the sulfite species [59]. There was a weak characteristic band assignable to the sulfite species at the noble metal sites in the SO 2 -treated 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample, which explains why the catalytic activity of this sample is not significantly altered. However, there was an obvious band at 1115 cm −1 attributable to the sulfite species at the noble metal site in the SO 2 -treated 0.47Pt/V 2 O 5 -TiO 2 sample, and its catalytic activity was remarkably decreased as a result. In addition, the characteristic band at 1400 cm −1 increased in intensity, which was ascribable to the vibration mode of VOSO 4 after the sample was treated in the presence of SO 2 . The vibration band intensity of the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample was stronger than that of the 0.47Pt/V 2 O 5 -TiO 2 sample, indicating that a greater amount of VOSO 4 was generated on the former sample, which protects the active noble metal site from being poisoned by SO 2 [60]. Therefore, the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample possessed a better sulfur dioxideresistance ability. In situ DRIFTS experiments were conducted to elucidate the formation and changes of the surface species on the 0.47Pt/V 2 O 5 -TiO 2 and 0.46PdPt 2.10 /V 2 O 5 -TiO 2 samples during the toluene oxidation process at different temperatures or reaction times. The adsorption process was undertaken in a (1000 ppm toluene + 20 vol% O 2 + N 2 (balance)) mixture flow at different temperatures, and the as-obtained in situ DRIFTS spectra of the 0.47Pt/V 2 O 5 -TiO 2 and 0.46PdPt 2.10 /V 2 O 5 -TiO 2 samples are shown in Figure 9A,C, respectively. The characteristic absorption band at 3040 cm −1 was attributed to the C-H bond stretching vibration of the aromatic ring in toluene, and the one at 2935 cm −1 was assigned to the asymmetric stretching vibration of the C-H bond of methyl in toluene [61]. The characteristic absorption band at 1498 cm −1 was caused by the planar skeleton vibration of the benzene ring [62]. It can be seen from Figure 9A,C that the characteristic band gradually disappeared with the rise in temperature. This result indicates that toluene is completely oxidized. According to the assignment of the characteristic absorption bands [63][64][65], we detected signals of the surface adsorbed C=O (at 1642 cm −1 ), anhydride (at 1850 cm −1 ), COO-(at 1545 cm −1 ), carboxylic acid (at 1380 cm −1 ), H 2 O (at 3653 and 3836 cm −1 ), chemisorbed oxygen (at 1050 cm −1 ), and other minor species during the toluene oxidation process. These results demonstrate that p-methylbenzoquinone is formed during the toluene oxidation process, and anhydride, benzoic acid, and benzaldehyde are further generated with the extension of the reaction. Finally, CO 2 and H 2 O were produced with the rise in temperature.
In order to investigate the difference of adsorbed species on the sample surface during the toluene oxidation process in the presence of 50 ppm SO 2 , we first activated the 0.47Pt/V 2 O 5 -TiO 2 or 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample in an oxygen flow at 250 • C for 1 h, and we then recorded in situ DRIFTS spectra of the adsorbed species at different times and 270 • C. As shown in Figure 9B,D, the absorption band at 1160 cm −1 was attributed to the vibration of SO 4 2− , whereas the one at 970 cm −1 was assigned to the vibration of SO 3 2− [66]. As shown in Figure 9C, the intensity of the absorption band at 970 cm −1 of 0.47Pt/V 2 O 5 -TiO 2 was significantly higher than that of 0.46PdPt 2.10 /V 2 O 5 -TiO 2 , indicating that SO 2 is adsorbed on the active Pt site in the sample to form the sulfite species. With the progress of the reaction, we can also see that the characteristic bands (at 2935 and 3040 cm −1 ) owing to toluene are significantly enhanced, indicating that the introduction of SO 2 competes the adsorption with toluene at the active site, and the oxidation of toluene was decreased. However, there was no obvious change in the characteristic band of toluene ( Figure 9D), indicating that the toluene oxidation activity of the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample was not affected by the introduction of SO 2 . As revealed in the XPS and (C 7 H 8 + SO 2 )-TPD characterization, SO 2 was mainly adsorbed at the vanadium site to form sulfate on the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample after SO 2 treatment. In addition, the doping of Pt with Pd could also protect the active Pt site from being poisoned by SO 2 , making the 0.46PdPt 2.10 /V 2 O 5 -TiO 2 sample show better sulfur dioxide resistance.

Preparation of the V 2 O 5 -TiO 2 Support
The V 2 O 5 -TiO 2 composite support was prepared using the hydrothermal method. In a typical preparation, 2.5 g of V 2 O 5 powder was added to 5.0 g of oxalic acid dissolved in 40 mL of deionized (DI) water, and the as-obtained suspension was then stirred at 85 • C for 4 h. After being cooled to room temperature (RT), the resulting blue aqueous solution was added dropwise to 30 mL of an isopropanol solution under stirring for 2 h. The above mixture was poured into an autoclave for thermal treatment at 180 • C for 20 h. After the thermally treated mixture was washed with 48 mL of DI water and ethanol (DI water/ethanol volumetric ratio = 5:1), centrifuged, and dried in an oven at 80 • C for 8 h, the as-obtained dried solid was calcined in a muffle furnace at a ramp of 3 • C/min from RT to 350 • C and kept at this temperature for 2 h, and thus the V 2 O 5 was obtained. Then, 0.6 g of the as-obtained V 2  The preparation procedures for xPdPt y /V 2 O 5 -TiO 2 (x and y are the PdPt y loading and Pt/Pd molar ratio, respectively) were similar to those of the above-mentioned 0.43Pt/V 2 O 5 -TiO 2 or 0.49Pd/V 2 O 5 -TiO 2 sample. Typically, 1.20, 0.90, or 0.62 mL of H 2 PtCl 6 and 0.52, 0.90, or 1.43 mL of PdCl 2 aqueous solution (Pt and Pd/PVA mass ratio = 1.0:1.5 mg/mg; theoretical Pt/Pd molar ratio = 7:3, 1:1, and 3:7) were added to 2 mL of the PVA aqueous solution (2.0 g/L) in an ice-water bath, which was protected from being illuminated by light. After being magnetically stirred for 30 min, 1.4, 2.1, or 2.3 mL of NaBH 4 aqueous solution (2.0 g/L; PdPt y /NaBH 4 molar ratio = 1.0:5.0 mol/mol) was rapidly added to the above precursor aqueous solution under stirring. The other preparation steps were the same as those for the preparation of 0.47Pt/V 2 O 5 -TiO 2 or 0.39Pd/V 2 O 5 -TiO 2 . The ICP-AES results reveal that the actual loadings (x) of PdPt y in xPdPt y /V 2 O 5 -TiO 2 are 0.46, 0.41, and 0.49 wt%, and the actual Pt/Pd molar ratios (y) are 2.10, 0.85, and 0.44, respectively. Therefore, the as-obtained catalysts were denoted as 0.46PdPt 2.10 /V 2 O 5 -TiO 2 , 0.41PdPt 0.85 /V 2 O 5 -TiO 2 , and 0.49PdPt 0.44 /V 2 O 5 -TiO 2 , respectively. It should be noted that during the process of loading noble metal NPs on V 2 O 5 -TiO 2 , the washing treatment could lead to the loss of some precious metals, thus resulting in the difference between the nominal and actual metal loadings.

Catalytic Evaluation
Catalytic activities of the samples were measured in a continuous flow fixed-bed quartz tubular microreactor (i.d. = 6.0 mm) at atmospheric pressure. A Shimadzu gas chromatograph (GC-2014C) was used for online analysis of the concentrations of the gases in the outlet of the microreactor. We loaded 25 mg of the sample well mixed with 125 mg of quartz sand (particle size: 40-60 mesh) in the microreactor. An O 2 flow of 20 mL/min was introduced into the microreactor, and the sample was pretreated at 250 • C for 1 h. The reactant gas mixture was composed of (1000 ppm toluene + 20 vol% O 2 + N 2 (balance)), and the space velocity (SV) was 40,000 mL/(g h). In the water, carbon dioxide, or sulfur dioxide resistance test, 5.0 vol% water vapor, 5.0 vol% CO 2 , or 50 ppm SO 2 (N 2 as the balance gas) was introduced to the reaction system over the typical sample at 245 • C. We adopted the temperatures at toluene conversions of 50% and 90% (T 50% and T 90% , respectively) to evaluate the catalytic activity of each sample. The specific reaction rate and turnover frequencies (TOF Pd , TOF Pt , or TOF Noble metal ) at a given temperature were also calculated. The detailed procedures for activity evaluation are described in the Supplementary Materials.