Cooperative Catalysis of Methane Oxidation through Modulating the Stabilization of PdO and Electronic Properties over Ti-Doped Alumina-Supported Palladium Catalysts.

Poor low-temperature catalytic activity and durability are the main drawbacks of palladium-based catalysts for methane combustion. Herein, stable and active PdO particles are constructed by incorporating Ti into an alumina support, which makes the catalysts exhibit satisfactory methane combustion activity. The results of comprehensive characterization reveal that an appropriate amount of Ti doping induces the optimization of electron transfer and distribution, thus contributing to the construction and stabilization of active PdO lattices. The reactive oxygen mobility is improved and the optimal PdO/Pd0 combination is achieved, thanks to the amplified PdO-support interaction. In addition, the acid-base properties are regulated and Brønsted acid sites are generated by virtue of the adjustment of electronic properties, which facilitate stabilization of PdO as well. Hence, the Ti-containing catalyst exhibits superior activity for methane oxidation at low temperatures. Notably, the activity and cyclic performance of the catalyst can be further enhanced when undergoing long-term and isothermal heat treatment under the reactant stream and methane, and it demonstrates a high performance with 90% CH4 conversion at 340 °C.

Introduction

Natural gas, a clean and efficient fuel, is gradually displacing diesel or gasoline in vehicles.[1−3] Due to incomplete combustion, the tail gases emitted from natural gas vehicles always contain residual methane, which would cause a serious greenhouse effect.[4,5] Catalytic combustion of methane is an effective way to reduce methane emission and also an important issue with respect to environmental sustainability. Nevertheless, the temperature of tail gases is relatively low (300–500 °C) during the vehicle cold-start period; therefore, the activity and thermal stability of catalysts at low temperatures should be improved to reach a high combustion efficiency.[6−8]

Methane combustion catalysts mainly include noble metals,[9−13] perovskites,[14] and hexaaluminates.[15] Among them, noble metal catalysts, especially Pd-based catalysts, are one of the most active catalysts for methane combustion.[16,17] Their catalytic activity depends strongly on the nature of the support, oxidation state, and stability of palladium species as well as the interaction with the support.[18,19] Although it is demonstrated that the Pd/Al2O3 catalyst is active in the methane combustion reaction, maintaining low-temperature (<400 °C) activity and improving the stability and efficiency of PdO particles still remain challenging subjects.[20] So far, the addition of metals or nonmetals into a support has been widely accepted as a promising way to solve these problems. Recently, our research group successfully incorporated phosphorus into an alumina support, which changed the acidity of the support, adjusted the distribution of palladium species, and modified the reduction properties of catalysts, and thus enhanced the low-temperature activity and hydrothermal stability of catalysts for methane combustion.[21] Zou[22] presented a facile way to stabilize the performance of Pd/Al2O3 by producing a spinel NiAl2O4 interface, which promoted the distribution of PdO and further inhibited the aggregation of PdO nanoparticles during the reaction. In addition, Venezia et al.[23] modified the methane combustion activity of supported palladium catalysts by incorporating titanium dioxide into an SiO2 support, proposing that the Si–O–Ti bonds in the mixed oxides were responsible for the enhanced activity. As a reducible oxide and oxygen carrier, TiO2 can enhance the reducibility and oxygen mobility of the supported palladium oxide for methane activation.[24] However, the specific surface area of TiO2 is usually low; thus, we envisaged that the introduction of titanium dioxide into mesoporous γ-Al2O3 may construct a support with relatively high surface area and oxygen transfer ability simultaneously, and the composition and properties of the catalysts may be further optimized.

Active palladium species and the interaction with supports are well known to be crucial for the activity and stability of Pd-based catalysts. During the methane combustion reaction, the PdO ↔ Pd0 redox cycle follows the Mars–van Krevelen mechanism.[22,25] Therefore, the presence of PdO is necessary for enhanced catalytic activity. Willis et al.[10] designed a catalyst with uniform palladium nanocrystals and systematically described structure–property relationships, verifying that PdO is the most active phase in catalytic methane combustion. In addition, Huang et al.[26] demonstrated that both Pd0 and PdO were the active sites under the condition of rich methane combustion (air/fuel < 1), whereas PdO played a dominant role in lean methane combustion (air/fuel > 1). Osman et al.[27] reported a zeolite-supported palladium catalyst with strong Brønsted (B-) acidity induced by the addition of TiO2, enhancing the reoxidation of Pd0 and facilitating the activation of methane. Therefore, the presence of titanium dioxide and its regulation on the acid–base properties of catalysts significantly affect the redox performance of palladium species. Based on the above literature, a reasonable design of catalysts may contribute to stabilize the active PdO and strengthen the interaction with supports, which consequently inhibits the sintering of PdO particles.

Inspired by the above assumption, an ultrasonic-assisted sol–gel technology was proposed to facilely prepare Ti-doped mesoporous γ-Al2O3 carriers to construct highly active palladium catalysts for low-temperature methane combustion. The structure–activity relationship of the catalysts is systematically investigated: (i) charge distribution and transfer between the support and palladium species enhances the metal–support interaction, and thus stabilizes the PdO particles and optimizes the redox properties of the catalyst. (ii) The presence of titanium oxide promotes the exchange of active oxygen species between PdO and the support, which contributes to the stabilization and utilization of PdO particles as well. (iii) The acid–base properties of the catalysts are effectively tuned by doping Ti into the lattice of γ-Al2O3. Overall, the influence of synergistic effects of electronics, active PdO stability, and acid–base properties of catalysts on methane combustion performance is discussed in detail.

Results and Discussion

Catalytic Performance

To determine the composition of the reactant feed in the activity evaluation, the influence of the CH4/O2 molar ratio on the catalytic oxidation of CH4 was investigated over a Pd/15TA catalyst (Figure S1). In our research, the temperature at 90% CH4 conversion (T90) is used as a criterion for evaluating catalytic activity; the lower the temperature, the better the catalytic performance. It can be found that the T90 of the catalyst was 435 °C when the molar ratio of CH4/O2 is 1:5. Then, the CH4 oxidation activity was enhanced with the CH4/O2 ratio increasing from 1:5 to 1:7.5 (T90 = 425 °C) and even to 1:10 (T90 = 410 °C). The reason for such behavior may be that more of the metallic Pd species are oxidized into PdO with the increase of O2 concentration,[26] providing more active species for methane oxidation. Then the activity slightly decreases as the oxygen content further increases to 1:12.5, which may be due to the partial inactivation of Pd species when covered with excessive oxygen under higher O2 concentration conditions.[28] Therefore, the molar ratio of CH4/O2 of 1:10 was chosen to carry out the subsequent activity tests.

The composition–activity relation of Pd/xTA catalysts for CH4 combustion under dry conditions was investigated, and the results are presented in Figure a. The activity profiles demonstrate that the catalytic activity varies with the Ti content. Among them, the Pd/15TA catalyst exhibits the highest activity (T90 = 410 °C); the T90 value is decreased by 60 °C compared with that of the Pd/0TA catalyst (T90 = 470 °C). However, the activity is no longer improved when the Ti doping content is higher than 15 wt %. This implies that the catalyst with an appropriate addition of Ti exhibits the highest catalytic performance. After the first round of activity test, the Pd/15TA catalyst was chosen to further explore the stability test at 390 °C (the corresponding conversion of CH4 was about 80%). As presented in Figure b, after a 20 h stability test, the CH4 conversion was reduced only by 5%, indicating that the catalyst could maintain good stability during the long-term test. Subsequently, the Pd/15TA catalyst after the stability test was selected to investigate the cyclic activity (Figure c). As shown in Figure c/cycle 1-1, the catalyst is first pretreated with a methane atmosphere at 400 °C for 30 min by shutting off oxygen. After that, the catalyst exhibits a remarkably high activity and T90 is 360 °C, which is lower than that of the activity test (T90 = 410 °C). And then, the temperature is decreased to 250 °C, and the second round of the activity test is carried out (without CH4 pretreatment), the T90 of which is 370 °C (Figure c/cycle 1-2). By analogy, cycles 2 and 3 repeat the process of cycle 1, and after the third cycle test (cycle 3), the T90 of the catalyst is maintained at 350 °C, indicating the presentable cyclic activity of Pd/15TA.

Figure 1

(a) Activity evaluation of Pd/xTA catalysts, (b) stability test at 390 °C, and (c) cyclic activity test of the Pd/15TA catalyst under dry conditions. Reaction conditions: 1 vol % CH4, 10 vol % O2, and N2 equilibrium gas, gas hourly space velocity (GHSV) = 50 000 mL g–1 h–1.

To further investigate the low-temperature durability of Ti-containing catalysts, the Pd/15TA catalyst after cyclic measurements (Figure c, cycle 3-2) is chosen and tested at 290 °C for 10 h. As shown in Figure S2, the CH4 conversion is ∼28% and decreases slightly to ∼20%. Then, the temperature is increased and fixed at 310 °C for 60 min, and the CH4 conversion is enhanced and maintained at ∼40%. In the next cooling–heating procedure, the CH4 conversion still remains unchanged at 290 and 310 °C, respectively. Subsequently, another cycle of activity evaluation is performed, and the T90 value further decreases to 340 °C. These results show that the catalyst possess satisfactory low-temperature stability and cyclic activity in the methane combustion reaction.

Another factor affecting the methane combustion of palladium-based catalysts is the presence of water vapor. Hence, the influence of H2O on the catalytic activity is evaluated in the presence of 5 vol % of H2O in the feed. The results shown in Figure a indicate that the T90 values of all catalysts are higher than those under dry conditions (Figure a), so we can conclude that the existence of water has a negative influence on the catalytic activity of all catalysts, which agrees well with that previously reported.[7] The reason for this kind of inhibition by H2O might be the formation of inactive Pd(OH)2 from PdO, significantly blocking the contact of CH4 with the active PdO species, and thus lowering the activity of the catalyst.[29] On the other hand, as compared with the undoped counterpart (T90 = 490 °C), all Ti-containing catalysts exhibit lower T90 values and higher catalytic activity, which show the same trend as those under dry conditions (Figure a).

Figure 2

(a) Methane conversion as a function of temperature over Pd/xTA catalysts and (b) cyclic stability test of the Pd/15TA catalyst under water conditions. Reaction conditions: 1 vol % CH4, 10 vol % O2, and 5 vol % H2O and N2 equilibrium gas, GHSV = 50 000 mL g–1 h–1.

To further investigate the hydrothermal stability, the used Pd/15TA catalyst after the cyclic measurement under dry conditions is used and evaluated at 350 °C in the presence of 5 vol % H2O (Figure S3). The CH4 conversion decreases from 87 to 71% during the 90 min stability test. This may be due to the fact that at the designated temperature (350 °C) and humidity (5 vol % H2O), water vapor persists and covers the active palladium site, blocking the contact of the reaction gas with the active site.[29] Thus, there are grounds to foresee that the activity will continue to decrease as the test time is further prolonged. It is worth noting that after the hydrothermal stability test the catalytic activity of Pd/15TA is retained (T90 = 390 °C) in four successive cycles (Figure b), indicating the superior cycling stability under wet conditions. Therefore, even though the hydrothermal stability of the catalyst needs to be improved, the above-mentioned results still manifest the high catalytic activity, as well as good cyclic and long-term stability (under dry conditions) of the Ti-containing catalyst (Pd/15TA), which demonstrate its competitive application potential for methane combustion in comparison to the reported palladium-based catalysts (Table S1).

Structural Properties of Catalysts

N2 Physisorption Measurements

Figure S4 represents the nitrogen adsorption–desorption isotherms of Pd/xTA catalysts, showing a typical IV isotherm with a H1 hysteresis loop, which implies a uniform mesoporous structure.[30] The loops for the samples with Ti content exceeding 15 wt % shift to a higher relative pressure, indicating a significant increase in the pore size,[31] as shown in Table S2. Besides, compared with the undoped one, the catalyst doped with 5 wt % Ti has a slightly higher Brunauer–Emmett–Teller (BET) surface area (SBET = 225 m2 g–1), whereas the SBET decreases when the Ti content exceeds 5 wt %. Considering that the specific surface area of the catalysts after Ti doping is reduced, and the activity is enhanced compared to that of Pd/0TA, we may reasonably conclude that the specific surface area and pore distribution are not the main influences on catalytic performance.

X-ray Diffraction (XRD) Patterns

XRD patterns of the Pd/xTA catalysts before and after the activity test are shown in Figure . The diffraction peaks at 2θ = 20.2, 32.3, 37.1, 39.4, 45.8, 60.7, and 66.8° are assigned to the (111), (220), (311), (222), (400), (511), and (440) reflections of γ-Al2O3 (JCPDS 10-0425), respectively. For Ti-containing catalysts (10–25 wt %), the peaks at 2θ = 27.3, 36.1, 54.3, 56.4, and 68.8° are attributed to the rutile TiO2.[32] And, the characteristic peaks at 2θ = 33.8 and 55° are ascribed to PdO.[33] For the fresh catalysts, there are no PdO peaks detected in Pd/0TA and Pd/5TA; however, the two peaks appear apparently after the activity test due to the occurrence of structural rearrangements.[8] Notably, when the Ti doping exceeds 10%, the PdO peaks of fresh Pd/xTA (x = 15–25) catalysts are detected and remain unchanged after the activity testing. Moreover, the peak intensity of PdO on the Pd/15TA catalyst after cyclic tests remains unchanged as well (Figure d, purple line), confirming the satisfactory stability of PdO particles. Combined with the results of activity tests, it can be concluded that the appropriate doping amount of Ti results in more PdO species in the catalyst, with higher dispersion and stability.

Figure 3

Wide-angle XRD patterns of Pd/xTA catalysts before (red line) and after (blue line) the activity test ((a) Pd/0TA, (b) Pd/5TA, (c) Pd/10TA, (d) Pd/15TA, (e) Pd/20TA, and (f) Pd/25TA).

Transmission Electron Microscopy (TEM) Analysis of Catalyst

Based on the highest catalytic activity for methane combustion, the Pd/15TA catalyst was chosen to conduct the high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) and energy-dispersive X-ray spectroscopy (EDS), which obviously illustrate the phase distribution. As shown in Figure a,b, the PdO particles homogeneously distribute over the Ti-doped support, and the mass fraction of palladium on the selected area is about 0.51 wt % (Figure b). EDS maps of Al, O, Ti, and Pd of the catalyst are shown in Figure c and further demonstrate the phase distribution in the Pd/15TA catalyst. The HRTEM images shown in Figure d–f provide more intuitive evidence for the size of Pd particles. It can be seen that the Pd species are distributed on the carrier with a size of 3–5 nm. Besides, interplanar distances of both PdO(110) plane (d = 0.26 nm) and (101) plane (d = 0.21 nm) can also be observed from Figures S5 and 4e, respectively, indicating the presence of PdO.

Figure 4

(a, b) HAADF-STEM images, (c) EDS elemental maps, and (d–f) HRTEM images of the Pd/15TA catalyst.

Chemical and Electronic States

O2-Temperature-Programmed Oxidation (TPO) and H2-Temperature-Programmed Reduction (TPR) Tests

O2-TPO profiles of Pd/xTA catalysts are presented in Figure a and show the peaks of O2-release owing to PdO decomposition during the heating process and O2-uptake from the reoxidation of Pd0 during the cooling process. In comparison with other catalysts, the Pd/15TA catalyst shows the highest PdO decomposition temperature during the heating process, which demonstrates the best thermal stability of the PdO particles and is responsible for the highest activity of Pd/15TA.[34,35] When the doping amount of Ti exceeds 15 wt %, in addition to the peak of oxygen desorption centered at 673.8 and 683.7 °C for Pd/20TA and Pd/25TA, respectively, there is a weak peak at about 740 °C appearing in both samples, which indicates that the oxygen adsorption states are obviously affected by the doping amount of Ti. Upon comparing the decomposition peak areas of O2-release shown in Table S3, the result manifests that the amount of oxygen released increases with increasing Ti doping amount up to 15 wt %, resulting from the increase in PdO content, thus facilitating the participation of PdO species in the methane oxidation reaction eventually. Furthermore, the temperature of O2-uptake peaks formed from Pd0 reoxidation shifts to a higher position with the increase of Ti content, that is, Ti doping reduces the difference between the decomposition temperature of PdO and the reoxidation temperature of Pd0 over the catalysts, especially over Pd/xTA (x = 10–25). It can be reasonably concluded that the redox cycle between PdO and Pd0 can easily proceed with the increase of Ti doping and that the addition of an appropriate amount of Ti ensures a suitable redox cycle. This can be attributed to the enhanced oxygen mobility induced by the addition of titanium oxide,[24] which facilitates the redox cycle between PdO and Pd0, realizing the conversion of CH4.

Figure 5

(a) O2-TPO and (b) H2-TPR profiles of Pd/xTA catalysts.

Normally, the PdO species are regarded as the active sites;[7,8,21] however, metallic Pd is reported to play important roles in the methane catalytic combustion as well.[36,37] So, H2-TPR was conducted to elucidate the effect of the PdO ↔ Pd redox cycle on the performance of catalysts. As shown in Figure b, the Pd/0TA catalyst shows that the reduction peak of finely dispersed PdO is located at 130 °C, which is consistent with the results reported in the literature.[38] The peak appearing at 410 °C is attributed to the reduction of subsurface palladium species,[39] and the broad peak located at 800–900 °C is due to the reduction of bulk oxygen.[40] Notably, the Pd/15TA catalyst exhibits the lowest reduction temperature of PdO (ca. 75 °C) relative to other catalysts, indicating that the addition of an appropriate amount of Ti could facilitate the reduction of PdO owing to the strong interaction between PdO and the Ti-doped support.[41] This strong interaction contributes to the migration of reactive oxygen species between the palladium species and the support, and the best redox ability of palladium is conducive to the combustion of methane. So, we suggest that the favorable PdO ↔ Pd redox cycle contributes to catalytic methane combustion. In addition, the PdO reduction temperature of the catalysts increases after excessive Ti doping. Combined with O2-TPO profiles, it can be concluded that excessive Ti doping is detrimental to the redox performance of the catalysts and the stability of PdO. Therefore, the activity of the catalysts is no longer improved.

X-ray Photoelectron Spectra (XPS) Analysis

Figure S6a shows the XPS spectra of Ti 2p at 458.6 and 464.4 eV corresponding to Ti 2p3/2 and Ti 2p1/2, respectively.[42] For the Ti-doped catalysts, the binding energy (BE) of Ti 2p3/2 gradually shifts to a lower position with increasing Ti content. This phenomenon can be related to the incorporation of some Ti species into the matrix of Al2O3 and then the formation of Ti–O–Al bonds.[42] And the doping of Ti leads to an increase in negative charge in the vicinity of Al3+ because of the excess oxygen around it, thus decreasing the binding energy of Al 2p (Figure S6b). The chemical states of palladium on the surface of Pd/xTA catalysts are presented in Figure a. The Pd 3d5/2 BEs within the range of 337.3–337.7 eV are ascribed to Pd2+, whereas the BEs at 335.9–336.3 eV are assigned to Pd0.[43,44] From Table , the doping of Ti increases the proportion of PdO in the catalysts. And the Pd/15TA catalyst shows satisfactory catalytic activity, maybe due to the suitable Pd2+/Pd0 ratio (Pd2+/Pd0 = 1.22).

Figure 6

(a) Pd 3d and (b) O 1s XPS spectra of Pd/xTA catalysts.

Table 1

XPS Analyses of Pd 3d5/2 and O 1s for Pd/xTA Catalysts

	relative content (%)
catalysts	Pd0	Pd2+	Pd2+/Pd0	Oads/OOH
Pd/0TA	55.4	44.6	0.81	1.18
Pd/5TA	49.2	50.8	1.03	1.31
Pd/10TA	47.2	52.8	1.12	1.46
Pd/15TA	45.1	54.9	1.22	1.66
Pd/15TA-used	46.1	53.9	1.17
Pd/20TA	44.1	55.9	1.27	1.55
Pd/25TA	43.0	57.0	1.33	1.27

In addition to the chemical states of the catalyst, the distribution of oxygen species in the catalyst is also an important factor affecting the catalytic activity. The BEs at 529.7–530.0 eV are ascribed to the lattice oxygen (Olatt) of the catalysts (Figure b). The BEs at 530.7–531.0 eV belong to surface-adsorbed oxygen (Oads), and that at 531.7–532.3 eV are assigned to hydroxyl groups and adsorbed water molecules (OOH).[41] In general, the reactivity of oxygen atoms on the surface of a catalyst is related to the thermodynamic affinity to the H atoms. Notably, Oads exhibits better affinity to H and higher mobility,[16,45] which is more effective for H-extraction during the C–H activation process. In contrast, OOH groups block the oxygen migration and exchange between PdO and the support by transforming PdO into inactive Pd(OH)2, resulting in the deactivation of catalysts. As shown in Table , there are more Oads and fewer OOH species in the Pd/15TA catalyst. However, excessive Ti doping can induce more OOH, which is not beneficial to catalytic methane combustion. As shown in Figure S7 and Table , compared with the fresh one, the Pd/15TA catalyst after the activity test (Pd/15TA-used) still maintains the appropriate Pd2+/Pd0 ratio (Pd2+/Pd0 = 1.17). This evidences that the active palladium species (PdO/Pd0) maintain a good redox cycle due to the presence of TiO2 as an oxygen carrier during the reaction process. As a result, an optimal distribution combination of PdO and Pd0 may benefit the activity and cyclic performance of the catalyst.

CO-Diffuse Reflectance Infrared Fourier Transform Spectroscopy (DRIFTS) and UV–vis Diffuse Reflectance (DR) Spectroscopy

In the methane combustion reaction, CO is regarded as a key intermediate affecting the reaction process. So, adsorption of CO on Pd/xTA catalysts was studied by in situ DRIFTS. As shown in Figure a, a strong band at 1917 cm–1 is assigned to bridge-bonded CO on the Pd(111) plane.[46] Compared with that of Pd/0TA, the band at 1917 cm–1 of Pd/xTA (x = 5–15) is located at a higher position, especially for the Pd/15TA catalyst, resulting from the weak adsorption of CO because the adjustment of the electronic structure and chemical states of the palladium species contribute to a strong interaction between the palladium species and the support, as concluded from the XPS spectra. The other two bands at 2083 and 2120 cm–1 are attributed to atop-bonded CO on the Pd0(111) defects and linear-bonded CO on ionic Pd+, respectively.[46] Particularly, the band at 2170 cm–1 is ascribed to CO adsorbed on Pd2+ ions, which is much weaker than the one adsorbed on Pd0, indicating the weak adsorption of CO on Pd2+.[47]

Figure 7

In situ CO-DRIFTS (a) and UV–vis DR (b) spectra of Pd/xTA catalysts before the reaction; and (c) UV–vis DR spectra of Pd/15TA, Pd/15TA-used (after reaction), and Pd/15TA-CH4 (pretreatment with CH4).

Additional details about the electronic properties of Pd/xTA are revealed by UV–vis DR spectroscopy, as shown in Figure b,c. For the Pd/0TA catalyst, there is no strong absorption observed in the available spectral region (λ > 200 nm) (Figure b), which agrees well with the strongly insulating character of γ-Al2O3.[48] As for Pd/xTA (x = 5–25) catalysts, there are broad absorption bands at 200–400 nm, which are caused by the charge transfer transition from O2– to Ti4+ (four-coordinated), that is, the excitation of electrons from the valence band (O 2p characteristics) to the conduction band (Ti 3d characteristics).[49] It worth noting that the enhanced band intensity may be due to the superimposition of charge transfer transitions of O2– → Ti4+ and Pd ↔ O. To gain further insights into the relationship between the structure and properties of catalysts, UV–vis DR experiment is conducted on Pd/15TA, Pd/15TA-used (after reaction), and Pd/15TA-CH4 (pretreatment with CH4) (Figure c). Compared with Pd/15TA, a red shift of the band is observed in Pd/15TA-CH4, which results from the change in dielectric constant around the metal nanoparticles.[50] And an enhanced intensity of the band is also detected; this may be because the pretreatment with CH4 increases the ability of charge transfer transition between O2– → Ti4+ and Pd ↔ O. Besides, the intensity of the absorption band of the Pd/15TA-used catalyst remains relatively strong, which further validates the result of XPS profiles of fresh and used catalysts and that the ratio of Pd2+/Pd0 tends to be stable (Figure S7 and Table ). Therefore, the nature of the interaction between active palladium and supports is electronic and involves charge transfer between active palladium and adjacent oxides. The above-mentioned results explain the improved catalytic activity after methane treatment and the high activity during cyclic reactions.

Acid–Base Properties

NH3- and CO2-Temperature-Programmed Desorption (TPD) Profiles

Figure shows the deconvolution of NH3-TPD and CO2-TPD profiles of Pd/xTA catalysts. Specifically, all patterns show three NH3 desorption peaks (Figure a), corresponding to weak (100–200 °C), medium (200–400 °C), and strong (400–700 °C) acid sites.[51] CO2-TPD profiles of the catalysts are presented in Figure b. The peaks at 30–150, 250–550, and 550–800 °C are also ascribed to weak, medium, and strong base sites, respectively.[18] The amount of acid and base sites is presented in Table . In comparison with that of the undoped sample, the total amount of acid and base sites of Ti-containing catalysts decreases significantly, especially for the Pd/15TA catalyst with the highest catalytic activity; in other words, a low amount of acid sites is beneficial to the catalytic methane combustion, which is in line with the result reported by Kinnunen et al.[52] And doping with excessive Ti increases the amount of acid sites, which is not good for methane oxidation. As evidenced by the higher desorption temperature of NH3, the strength of medium or strong acid sites of Ti-modified catalysts is enhanced, which is attributable to the formation of bridged hetero Ti–O–Al bonds resulting in excessive charges.[53] It is also observed that the desorption peaks of CO2 on the strong base sites shift to a lower temperature with increasing Ti content, especially for the Pd/15TA catalyst, demonstrating the weak adsorption capacity for CO2.[38]

Figure 8

(a) NH3-TPD and (b) CO2-TPD patterns of Pd/xTA catalysts.

Table 2

Acid/Base Sites and Distribution of Pd/xTA Catalysts Derived from NH3-TPD and CO2-TPD

	acid sites (mL g–1)				base sites (mL g–1)
catalysts	weak	medium	strong	total	weak	medium	strong	total
Pd/0TA	0.4	0.83	2.14	3.37	0.007	0.061	0.132	0.20
Pd/5TA	0.38	0.84	1.64	2.86	0.003	0.042	0.065	0.11
Pd/10TA	0.32	1.24	1.14	2.70	0.002	0.036	0.062	0.10
Pd/15TA	0.22	1.01	0.81	2.04	0.001	0.022	0.037	0.06
Pd/20TA	0.18	1.15	1.10	2.43	0.001	0.038	0.051	0.09
Pd/25TA	0.32	1.21	1.48	3.01	0.001	0.034	0.035	0.07

Pyridine-Infrared (Py-IR) Analysis

To further investigate the influence of acid type of Pd/xTA catalysts on the catalytic performance, a Py-IR analysis was performed (Figure ). The peaks at 1614 and 1448 cm–1 are ascribed to pyridine coordinated with strong Lewis (L)-acid sites and 1576 cm–1 corresponds to weak L-acid sites.[54,55] Besides, the peak at 1491 cm–1 is also related to pyridine coordinated with L-acid sites, as well as the possible contributions from protonated pyridinium species on B-acid sites.[56] To elucidate the effect of Ti doping on the amount of L-acid sites, we compared the peak intensity of L-acid sites (1614 and 1448 cm–1) of the Pd/0TA catalyst with those of Pd/15TA, which show that titanium doping reduced the amount of L-acid sites. Notably, only Ti-doped catalysts show weak peaks at 1644 and 1541 cm–1, which are ascribed to pyridine adsorbed on B-acid sites.[56] Combined with the results of NH3-TPD and XPS, the introduction of Ti leads to excess negative charges, which are probably balanced by protons adsorbed on the surface, eventually causing the formation of B-acid sites.[57] It is established that the presence of B-acid sites would make the palladium species electron-deficient, thereby promoting the oxidation of Pd0 and leading to the enhancement of CH4 combustion activity.[27] Moreover, the adsorption band at 1644 cm–1 of the Pd/15TA catalyst is stronger than that of the other catalysts, suggesting that an appropriate doping content of Ti generates more Brønsted-acid sites.

Figure 9

Pyridine-IR spectra of Pd/xTA catalysts desorbed at 200 °C.

Discussion on a Possible Promotion Effect

Methane oxidation over a Pd-based catalyst is a complex reaction, which may involve the activation of C–H on active sites in the vicinity of both Pd and PdO.[16,58] Generally, PdO is the active phase for methane complete combustion, and an appropriate proportion of Pd2+/Pd0 is beneficial to catalytic methane combustion (Figure ). Therefore, the stability of PdO leads to the enhancement of methane oxidation. As can be seen from O2-TPD profiles (Figure S8), compared with that of Pd/0TA, the temperature of PdO decomposition of Pd/15TA shifts to a higher position, indicating that the doping of Ti increases the interaction of PdO particles with the support, which strongly confirms the results of O2-TPO and XRD patterns.

Based on the above analyses, we propose a possible promotion effect of Ti-containing catalysts for catalytic methane combustion. XPS spectra of Ti 2p and NH3-TPD profiles of Pd/xTA catalysts affirm that Ti incorporates into the lattice of alumina and displaces the Al3+ sites to form Ti–O–Al bonds, which increase the negative charge around the Al3+ ions. As a result, the interaction between PdO particles and the support is enhanced, leading to the improvement of the thermal stability of active PdO, which has been confirmed by XRD patterns and O2-TPO spectra. Notably, the excessive negative charge can be balanced by the surface protons, accompanied by the production of Brønsted-acid sites (Py-IR spectra), which can induce an electron deficiency in palladium species, facilitating the oxidation of Pd0 and improving CH4 combustion activity.

Besides, the CH4-TPD was performed to study the adsorption behavior of CH4 molecule on the Ti-containing catalyst. As illustrated in Figure S9, the desorption peaks of CH4 are observed at 380 and 604 °C for the catalyst, and the desorption peak attributable to CH3 is detected at 278 °C. It is indicated that methane is activated and dehydrogenized to generate CH3 and H during methane adsorption (CH4 → CH3 + H2 (or H)), and the CH3 group will be adsorbed on the surface of the catalyst, which is consistent with the proposed mechanism by Fujimoto et al.[59] that the rate-determining step of CH4 combustion involves the dissociation adsorption of CH4 on a Pd surface vacancy and Pd–O species site pair.

In addition, the desorption peaks of H2O are observed in the illustration at 376 and 592 °C (Figure S9). Based on the mechanism described by Fujimoto,[59] we assume that the separated H atom from CH4 would combine with Pd–O bonds to form Pd–OH species, and H2O is generated by quasi-equilibrated condensation of Pd–OH species. Therefore, the desorption peaks of H2O can be observed. Remarkably, a CH4 molecule may be activated by the Brønsted acid sites and then be adsorbed by oxygen (Oads), or it may react with Pd oxides (CH4 + 4O(PdO) → CO2 + 2H2O) to create CO2 and H2O.[4,60] So, the mass signal of CO2 is observed. In particular, this process is effectively promoted due to the higher proportion of Oads in the Pd/15TA catalyst, as evidenced by the XPS spectra of O 1s (Figure b). Moreover, the presence of a small amount of TiO2 on the catalyst surface can improve the migration ability of oxygen species, especially Oads, and subsequently, the vacancy of Oads will be refilled with the feed gas oxygen.[61]

Conclusions

Through an ultrasonic-assisted sol–gel method, Ti-doped alumina was facilely synthesized and taken as supports for palladium catalysts for methane combustion. Results demonstrate that the incorporation of Ti adjusts the charge distribution, which induces the generation of stable PdO particles and the enhancement of PdO–support interactions. Meanwhile, the migration of reactive oxygen species can be effectively facilitated due to the existence of oxygen carrier TiO2, and the tailored performance contributes to the redox properties and regenerability of PdO particles. In addition, the acid–base properties over Ti-modified catalysts are modulated as well. Moreover, the strategy of long-term and isothermal heat treatment under the reactant stream effectively improves the catalytic activity, and the catalyst possesses cyclic stability. The above findings highlight the potential of electronic and acid–base properties in stabilizing PdO particles, which can be extended to fabricate other high-performance and sustainable palladium catalysts.

Experimental Section

Support Preparation

Ti-doped mesoporous alumina supports were synthesized by an ultrasonic-assisted sol–gel method: 10 mmol aluminum isopropoxide was first dispersed in 10 mL of isopropanol by ultrasound to form solution A. Then, 0.17 mmol P123 (EO20PO70EO20, Mav = 5800) was dissolved in a mixed solution of absolute ethanol (10 mL) and glacial acetic acid (0.1 mL) under ultrasonic conditions to form solution B, accompanied by the addition of a required amount of isopropyl titanate solution (C12H28O4Ti, 95%). Subsequently, the two solutions were mixed, and then the ultrasonic reaction was continued for 1.5 h. The resultant milky white solution was aged at room temperature for 10 h and then dried in an 80 °C drying oven overnight. The resulting samples were calcined at 500 °C (1 °C min–1) and held for 4 h, and then calcined at 900 °C (10 °C min–1) for 1 h. The final samples were denoted as xTA, and the mass fraction of Ti (x = 0, 5, 10, 15, 20, and 25) was denoted as x, which was calculated by the equation as follows:where mTi and mAl represent the mass of Ti and alumina, respectively.

Catalyst Preparation

Catalysts with 0.5 wt % Pd supported on xTA were prepared by the incipient wetness impregnation method, and Pd(NO3)2 aqueous solution (0.025 g mL–1) was chosen as the metal precursor. Typically, the xTA support was immersed in an aqueous Pd(NO3)2 solution at room temperature and held overnight. The as-impregnated samples were dried at 100 °C for 10 h and then calcined at 500 °C for 1 h in air with a heating rate of 10 °C min–1. The resultant samples were marked Pd/xTA.

Characterization Techniques

N2 physisorption measurements were carried out at 77 K on an ASAP 2460 apparatus. The Brunauer–Emmett–Teller (BET) method was utilized to calculate the specific surface areas (SBET). The pore volume and pore size distributions were derived from the desorption branches of isotherms using the Barrett–Joyner–Halenda model. Powder X-ray diffraction (XRD) was measured on a Philips X’Pert Pro MPD diffractometer using Cu Kα radiation (λ = 0.15406 nm). Transmission electron microscopy (TEM) images, high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) images, and energy-dispersive X-ray spectroscopy (EDS) images were obtained using a Talos F200X transmission electron microscope with an accelerating voltage of 200 kV.

Temperature-programmed oxidation (TPO) and temperature-programmed reduction (TPR) experiments were conducted on a quartz tube microreactor with a thermal conductivity detector (TCD). O2-TPO: The catalyst powder (100 mg) was pretreated prior to the measurement under a flow of He (30 mL min–1) for 30 min at 300 °C. Subsequently, the catalyst was exposed to a mixture of 2 vol % O2 in He (50 mL min–1), and the temperature was then increased to 900 °C (10 °C min–1) and cooled down. Oxygen release–uptake was evaluated using a GC-14C gas chromatograph with a TCD detector. H2-TPR: Before the measurement, the catalyst (100 mg) was precleaned in N2 (30 mL min–1) holding at 300 °C for 30 min and was then cooled to room temperature. The data were collected from room temperature to 800 °C under 10 vol % H2/Ar (30 mL min–1) at a heating rate of 10 °C min–1. The results were obtained using a GC-14C gas chromatograph with a TCD.

X-ray photoelectron spectra (XPS) were obtained on an ESCALAB 250 multifunctional electronic energy spectrometer using Al Kα (E = 1486.6 eV) as the X-ray source. Binding energies were calculated on the basis of C 1s at 284.8 eV.

In situ diffuse reflectance infrared Fourier transform spectra of CO adsorption (CO-DRIFTS) and Fourier transform infrared spectra of pyridine adsorption (Py-FTIR) over catalysts were recorded using a Nicolet 6700 spectrometer equipped with an MCT with a resolution of 2 cm–1. CO-DRIFTS: First, the catalyst was pretreated with 10 vol % H2/Ar at 300 °C for 30 min, and then cooled to room temperature under helium gas. Subsequently, CO adsorption experiment was carried out under 1 vol % CO/He (30 mL min–1) for 1 h. Py-FTIR: Prior to the measurement, the catalyst was pretreated at 300 °C for 30 min to remove moisture. Pyridine adsorption was conducted at 30 °C for 30 min and then evacuated at a designated temperature.

Ultraviolet–visible (UV–vis) spectra were recorded on a PerkinElmer Lambda 950 instrument with a diffuse reflectance spectroscopy technique. All of the spectra were collected and converted to Kubelka–Munk units, and BaSO4 (AR) was used as a reflectance standard.

Temperature-programmed desorption of NH3 (NH3-TPD) and temperature-programmed desorption of CO2 (CO2-TPD) were conducted on an AutoChem 2910 instrument. In a typical process of NH3-TPD, 50 mg of the sample was pretreated at 300 °C for 30 min under a helium stream and then cooled to 30 °C. Then, 8.3 vol % NH3/He stream was introduced at room temperature and kept for 1 h (30 mL min–1). Desorption was carried out with the temperature increasing up to 800 °C (10 °C min–1), and the ammonia desorption was continuously monitored using a TCD. The test process of CO2-TPD was similar to that of NH3-TPD. The adsorbed gas was 100 vol % CO2 and the amount of sample was 200 mg.

Temperature-programmed desorption of CH4 (CH4-TPD) was carried out in a quartz microreactor with a DYCOR DEM200MS mass spectrometer detector. Typically, 100 mg of the sample was adsorbed by 10 vol % CH4/Ar (30 mL min–1) at 400 °C for 30 min and then cooled to 30 °C. Desorption was carried out with the temperature increasing from 30 °C up to 800 °C under a helium stream. The composition of the products was detected by online mass spectrometry with m/z ratios of 44 (CO2), 28 (CO), 18 (H2O), 16 (CH4), and 15 (CH3).

Temperature-programmed desorption of O2 (O2-TPD) experiments were performed using the same reaction apparatus as for O2-TPO. First, the catalyst (100 mg) was pretreated at 300 °C under He with a flow of 30 mL min–1 for 30 min and then exposed to 2 vol % O2/He (50 mL min–1) at 400 °C for 60 min. Subsequently, the catalyst was cooled to room temperature. After the catalyst was purged by He for another 30 min, the catalyst was heated to 950 °C (10 °C min–1). The O2 signals were detected by a GC-14C gas chromatograph with a TCD.

Catalytic Measurement

Activity measurement for methane combustion was carried out in a fixed-bed quartz flow reactor (shown in Figure S10), and the dosage of catalysts was 0.1 g. The reaction gas containing 1 vol % CH4 (4 vol % CH4 in N2), 10 vol % O2 (21 vol % O2 in N2), and N2 equilibrium gas was supplied at a gas hourly space velocity (GHSV) of 50 000 mL g–1 h–1. For the performance test under wet conditions, 5 vol % water vapor was added to the feed stream. The products of the reaction were periodically analyzed online by an Agilent technologies 7820A GC system. The CH4 conversion was denoted as X and calculated by the equation as follows:where [CH4]in is the inlet flow of CH4 and [CH4]out is the outlet flow of CH4.