Promotion of TiO2 Nanotube-Confined Pt Nanoparticles via Surface Modification with Fe2O3 for Ethylene Oxidation at Low Temperature.

A modified confined catalyst with Pt nanoparticles on the interior and Fe2O3 on the exterior surface of TiO2 nanotubes (Pt-in/Fe2O3-TNTs) was prepared and investigated for catalyzing the oxidation of ethylene. Compared with the Pt-in/TNTs without Fe2O3 modification, the Pt-in/Fe2O3-TNTs exhibited a significantly enhanced activity, and the complete conversion temperature of ethylene decreased from 170 to 95 °C. X-ray photoelectron spectroscopy analysis indicated that the Pt nanoparticles were stabilized at higher oxidation states in the Pt-in/Fe2O3-TNT catalyst. It was proposed that the modification of Fe2O3 on the outer surface can tune the electronic state of the encapsulated Pt particles and accelerate the electrons transferred from Pt to Fe species via TiO2 nanotubes, thus improving the catalytic oxidation performance of the confined catalyst.

Introduction

Ethylene, a low-molecular-weight gaseous volatile organic compound (VOC), is widely used as a raw material in chemical industry, which will lead to photochemical pollution of the atmosphere and cause anesthetic illness.[1] Moreover, ethylene released from fruits and vegetables can also accelerate the maturation and decay of produce.[2,3] Therefore, the removal of ethylene contaminant from air under mild conditions is of great necessity. Low-temperature catalytic oxidation is one of the most important and effective methods for transforming ethylene to harmless carbon dioxide and water.

Until now, some solid thermal catalysts have been attempted for the oxidation of ethylene, such as Au/Co3O4,[4] Pt/MnO2,[5] Cu/MnO2,[6] Ag/ZSM-5,[7,8] Pd/ZSM-5,[9] Pt/ZSM-5,[10] Pt/MCM-41,[11] Pt/SBA-15,[12,13] Pt/C,[14] Pt/YSZ,[15,16] Ru/YSZ,[16] Pt/ZrO2,[11] Pt/TiO2,[11] and Pt/SiO2.[17] Among these catalysts, supported Pt catalysts have attracted wide interest because of their superior activity in catalyzing the oxidation of ethylene at relatively low temperatures. It was reported that the Pt/MCM-41 showed higher activity to catalytically remove C2H4 than other noble metals (Pd, Au, or Ag) loaded on the same support.[11] Although the conversion of trace ethylene over Pt nanoparticles supported on zeolites and mesoporous silica reached 100% at 0 °C, the activity of these low-temperature catalysts decreased rapidly and did not meet the requirements for practical applications.[10−13] Pt catalysts with metal oxides (such as Al2O3, ZrO2, TiO2, YSZ, and MnO2) as supports were also widely studied for the oxidation of ethylene in view of the synergetic catalytic effect between noble metals and metal oxides.[5,11,15−17] However, most of these catalysts showed relatively low catalytic activities, and the temperature for the complete conversion of ethylene was usually higher than 100 °C. It is therefore still a great challenge to develop a catalytic material with enhanced low-temperature activity and excellent cycle stability toward the complete oxidation of ethylene.

The confinement of the active species inside a nanotube space has been proved as an efficient strategy to improve the catalytic performance. Bao research group[18−20] reported that the confinement within carbon nanotubes (CNTs) can strengthen the electrons transferred from the metal to the carbon surface and hinder the sintering of the encapsulated active component, thereby significantly enhancing the activity and stability of the catalysts. In our previous works, we have successfully entrapped noble metals into titanium dioxide nanotubes (TNTs) and found that TNTs exhibited a similar confinement effect as CNTs due to the unique tubular morphology.[21−23] The performances of phenol hydrogenation, catalytic combustion, and photocatalytic hydrogen production were notably enhanced when Pt, Pd, and Au nanoparticles were entrapped into TNTs.[21−23] It is known that the curvature of nanotube walls causes the electron density to shift from the concave inner to the convex outer surface, which makes the nanotube exist with an electron-deficient interior and electron-enriched exterior surface.[24,25] Regulating the electron properties of materials by introducing an electron-withdrawing group is another effective method to improve the catalyst performance.[17,26−29] Nanometer iron oxide (Fe2O3) was intensively studied in the fields of thermal-, electro-, and photocatalysis.[30−32] As a variable valence metal oxide with the property of binding electrons, Fe2O3 can easily capture electrons and get reduced into Fe3O4. Therefore, the electron property of the encapsulated active component can be further regulated by coupling Fe2O3 on the outside wall of TNTs. Our previous study has revealed that the electronic state of Pd is significantly enhanced when it is modified with Fe2O3 nanoparticles.[29] As a result, the modified catalysts presented a significantly improved activity for the photocatalytic degradation of methyl orange. The confined Pt nanoparticles are considered highly active and promising catalysts for the combustion of VOCs. The modulation on the electronic property of the Pt catalysts will provide some insights for the design of highly efficient catalytic combustion catalysts.

In the present work, a modified confined catalyst, with Pt nanoparticles inside and Fe2O3 on the outside surface of TNTs (Pt-in/Fe2O3-TNTs), was successfully fabricated using vacuum-assisted incipient-wetness impregnation and deposition–precipitation methods (Scheme ). The ethylene oxidation performance of Pt-in/Fe2O3-TNTs was compared with that of the confined Pt catalyst (Pt-in/TNTs) without Fe2O3 modification. The contribution of Fe2O3 modification toward the modulation of electron structure and the enhancement of catalytic oxidation activity was discussed in detail.

Scheme 1

Synthetic Route for the Pt-in/Fe2O3-TNT Catalyst

Results and Discussion

Figure shows the XRD patterns of the TNTs, Fe2O3-TNTs, Pt-in/TNTs, and Pt-in/Fe2O3-TNTs samples. All samples display the characteristic peaks of anatase TiO2 (JCPDS no. 21-1272). The relative weak peaks at 2θ = 39.8 and 46.2° can be differentiated in the Pt-in/TNTs and Pt-in/Fe2O3-TNTs catalysts, corresponding to the (111) and (002) planes of the metallic Pt crystalline phase (JCPDS no. 04-0802). It is observed from zooming in of the regions of 2θ = 30–50° that the diffraction peaks of metallic Pt are slightly weakened with the deposition of iron oxide on the catalyst surface. These results imply that an interaction between Pt particles and iron oxide may be present in the Pt-in/Fe2O3-TNT catalysts. No distinct diffraction peaks indexed to the Fe2O3 phase are observed in the XRD patterns of the Fe2O3-TNTs and Pt-in/Fe2O3-TNTs samples, which might be due to the hindrance by the strong TiO2 peaks or the formation of amorphous particles which are below the XRD detection limit. For comparison, pure Fe2O3 was obtained by the precipitation method using Fe(NO3)3·9H2O as the iron resource and Na2CO3 as the precipitation agent. For the pure Fe2O3 calcined at 250 °C under an air atmosphere, the weak diffraction peaks appear at 2θ = 24.1, 33.1, 35.6, 40.8, 49.4, and 54.0° (see Figure S1), which could be associated with the presence of the α-Fe2O3 phase.

Figure 1

(a) XRD patterns of the as-prepared catalysts and (b) the zooming in the regions of Pt peaks (2θ = 30–50°).

Figure shows the TEM and high-resolution TEM (HRTEM) images of pristine TNTs, Pt-in/TNTs, and Pt-in/5% Fe2O3-TNTs samples. It can be seen from Figure a–b that the pristine TNTs display one-dimensional nanotubular morphology with a relative smooth and clean outside surface. Also, the TNTs obtained are multilayered nanotubes with the outer diameter of 9–12 nm and the inner one of 5–8 nm. For the Pt-in/TNTs and the Pt-in/5%Fe2O3-TNTs, most of the Pt nanoparticles with the average size of 2–3 nm are successfully confined within the inner cavity of the TiO2 nanotube, as shown in Figure c–f. The HRTEM image (the inset of Figure d) recorded from the region marked with yellow circles shows a lattice spacing of 0.226 nm, which corresponds to the metal Pt (111) plane. The filling of Pt nanoparticles inside the nanotube can be attributed to the capillary effect derived from the vacuum-assisted impregnation procedure.[21−23] For the Pt-in/5% Fe2O3-TNTs, Fe2O3 particles are well dispersed on the external surface of TNTs, as depicted by the orange circles with a measured lattice spacing of 0.252 nm corresponding to the Fe2O3 (110) plane (Figure e–f). The particle size distributions of Pt from the TEM analysis are provided in the Supporting Information (see Figure S2).

Figure 2

TEM and HRTEM images of the prepared catalysts: (a,b) TNTs, (c,d) Pt-in/TNTs, and (e,f) Pt-in/5% Fe2O3-TNTs.

High-angle annular dark field (HAADF)-STEM images and STEM–EDX elemental mappings of the Pt-in/Fe2O3-TNTs were performed to further determine the distribution of elements. As shown in Figure , better distribution and coexistence of Ti, O, Pt, and Fe elements can be found for the Pt-in/Fe2O3-TNTs, implying that Pt and Fe2O3 nanoparticles are dispersed uniformly in/on the TNTs. STEM–EDX spectrum and quantitative results of the Pt, Fe, and Ti elements for the Pt-in/5% Fe2O3-TNT catalyst are provided in the Supporting Information (see Figure S3). The Pt/Ti/Fe atomic ratio is calculated to be 0.0059/1/0.0528. Compared with the value determined by inductively coupled plasma atomic emission spectroscopy (ICP-AES) (Pt/Ti/Fe atomic ratio = 0.0082/1/0.0346), the Fe signal is enhanced at the TiO2 edge, further confirming that Fe2O3 is deposited on the outer surface of the nanotubes.

Figure 3

HAADF-STEM images and STEM–EDX elemental mappings of the Pt-in/5% Fe2O3-TNTs: (a) base, (b) gray, (c) Ti, (d) Fe, (e) O, and (f) Pt.

Figure shows the N2 adsorption–desorption isotherms and pore size distribution curves of the synthesized samples. The corresponding surface area (SBET), pore volume (Vp), and average pore size (Dp) are summarized in Table . All samples exhibit typically a BDDT Type III isotherm with a large type H3 hysteresis hoop according to the Brunauer–Deming–Deming–Teller classification. The support of TNTs has a surface area of 151.4 m2/g and a pore volume of 0.79 cm3/g. The surface area, pore volume, and pore size decrease on loading Pt on TNTs. For the Pt-in/TNTs and Pt-in/Fe2O3-TNTs samples, the surface area and pore volume are clearly lower than those of the TNTs. This can be explained by the fact that Pt nanoparticles are confined within the inner cavity of TiO2 nanotubes via the vacuum-assisted impregnation process and take up some of the pore volume. The actual loading contents of Pt and Fe2O3 in the samples were determined by ICP-AES, as shown in Table . The actual loading of Pt is 2.01 wt % for the as-prepared catalysts, and the actual loading of Fe2O3 is 1.06 wt % for the Pt-in/1% Fe2O3-TNTs and 3.46 wt % for the Pt-in/5% Fe2O3-TNTs, respectively.

Figure 4

(a) N2 adsorption–desorption isotherms and (b) pore size distributions of the samples.

Table 1

Textural and Chemical Properties of the As-Prepared Samples

sample	SBET (m2/g)	Vpa (cm3/g)	Dpa (nm)	DPtb (nm)	Ptc (wt %)	Fe2O3c (wt %)	atomic ratioc Pt/Ti/Fe
TNTs	151.4	0.79	17.3
Pt-in/TNTs	120.4	0.74	16.9	2–3	2.01		0.0082/1
Pt-in/1% Fe2O3-TNTs	122.4	0.55	12.1	2–3	2.01	1.02	0.0082/1/0.0102
Pt-in/5% Fe2O3-TNTs	132.1	0.67	12.3	2–3	2.01	3.46	0.0082/1/0.0346

Porosity data were derived by the Barrett–Joyner–Halenda desorption branch.

Average particle sizes were calculated from 200 individual crystallites in TEM images.

The actual loading contents of Pt and Fe2O3 and the atomic ratio of Pt/Ti/Fe were determined by ICP-AES.

X-ray photoelectron spectroscopy (XPS) measurements were conducted to investigate the chemical states of the surface elements in the samples. The photoelectron peaks corresponding to Fe 2p, O 1s, Ti 2p, C 1s, and Pt 4f can be detected in the full-survey-scan spectrum for the Fe2O3-loaded catalysts (see Figure S4), indicating that the Fe element has been successfully introduced in the synthesized samples. The high-resolution XPS spectra recorded are shown in Figure . For the Pt-in/1% Fe2O3-TNTs and Pt-in/5% Fe2O3-TNTs catalysts, the Fe 2p3/2 and Fe 2p1/2 peaks centered at 711.2 and 724.6 eV along with two shakeup satellite peaks at about 719.5 and 732.2 eV are in good agreement with the reported Fe2O3 phase (Figure a).[34,35] No obvious peaks from Fe0 and/or Fe2+ were observed, indicating that Fe2O3 is the only iron species formed on the surface of the Pt-in/Fe2O3-TNTs. The Pt 4f XPS spectra can be deconvoluted into two peaks, with the lower binding energy peak corresponding to the Pt 4f7/2 level and the higher binding energy peak corresponding to the Pt 4f5/2 level. For the Pt-in/TNTs, there are two Pt 4f7/2 peaks centered at 70.3 and 71.2 eV, which can be assigned to Ptδ− and Pt0 species, respectively.[28] Modification by Fe2O3 significantly changes the profiles of Pt 4f core electrons. Compared with the unmodified Pt-in/TNTs, the Pt 4f7/2 peaks for the Fe2O3-loaded catalysts shift to a higher binding energy value centered at about 71.3 and 72.3 eV (Figure b), corresponding to metallic Pt0 and ionic Ptδ+ species.[10] This implies that the charge is transferred from Pt to Fe2O3 species via TiO2 nanotubes, and Pt nanoparticles are stabilized at higher oxidation states in the Pt-in/Fe2O3-TNTs, indicating the formation of strong metal–support interactions between Pt and TNTs. As a result, the electron density of the Pt nanoparticles is affected by the modification of Fe2O3 nanoparticles on the outside wall of the TNTs. Moreover, the modification of Fe2O3 on the TNTs does not change the XPS spectra of Ti 2p. As shown in Figure c, the Ti 2p3/2 and Ti 2p1/2 centered at 458.9 and 464.7 eV are assigned to Ti4+ for all samples. The integral peak of O 1s can be divided into two peaks concentrated at approximately 530.1 and 532.2 eV (Figure d), corresponding to oxygen in the composition of titanium oxide and oxygen in the composition of OH group or water. The surface atomic ratios of the different Pt species (Ptδ+/Pt0/Ptδ−) and the Pt/Ti/Fe for all samples are listed in Table . The Ptδ+/(Pt0 + Ptδ+) ratios in the Pt-in/Fe2O3-TNT samples are significantly increased with the deposition of Fe2O3 on the outer surface of TNTs (25.6% for Pt-in/1% Fe2O3-TNTs vs 42.7% for Pt-in/5% Fe2O3-TNTs). As is known, the inner surface of TNTs is in an electron-deficient state, and such a state is further strengthened by introducing an electron-withdrawing group (Fe2O3) on the outer surface. Meantime, the position of the conduction band of Fe2O3 (0.28 eV) is lower than that of anatase TiO2 (−0.29 eV) in terms of energy level of the materials. The difference in conduction band positions will drive the electron to transfer from TNTs to Fe2O3 particles crossing the interface. Therefore, the encapsulated Pt species are stabilized at higher oxidation states in the Pt-in/Fe2O3-TNT samples, which would exert further influence on their oxidation activity. Furthermore, it also can be seen from the results of XPS and ICP-AES that the Fe/Ti atomic ratios on the surface for the Pt-in/Fe2O3-TNT catalysts are much larger than that in the bulk (see Tables and 2), indicating that Fe2O3 is highly dispersed on the outer surface of TNTs.

Figure 5

XPS spectra of the prepared catalysts: (a) Fe 2p, (b) Pt 4f, (c) Ti 2p, and (d) O 1s.

Table 2

XPS Results of the As-prepared Catalysts: Binding Energy and Element Percentage

	BE for Pt 4f7/2 (eV)			element percent (%)			atomic ratio
catalyst	Ptδ−	Pt0	Ptδ+	Ptδ−	Pt0	Ptδ+	Pt/Ti/Fe
Pt-in/TNTs	70.3	71.2		68.4	31.6		0.0075/1
Pt-in/1% Fe2O3-TNTs		71.2	72.2		74.4	25.6	0.0052/1/0.0316
Pt-in/5% Fe2O3-TNTs		71.3	72.3		57.3	42.7	0.0054/1/0.1631

The catalytic performances of Fe2O3-TNTs, Pt-in/TNTs, and Pt-in/Fe2O3-TNTs catalysts for ethylene oxidation were investigated under the flow of 2.0% C2H4/49% O2/Ar. Figure shows the curves of ethylene conversion over the different catalysts. The conversion of ethylene over the Fe2O3-TNTs is extremely low, suggesting that the oxidation of ethylene cannot proceed at all in the absence of Pt nanoparticles below 200 °C. The Pt-in/TNTs exhibit complete conversion of C2H4 to CO2 at 170 °C. After the modification by Fe2O3, the ethylene oxidation activities of the catalysts are obviously improved. The complete conversion temperature of ethylene decreases from 170 to 100 °C and 95 °C for the Pt-in/1% Fe2O3-TNTs and Pt-in/5% Fe2O3-TNTs, respectively. In comparison with the Pt-in/1% Fe2O3-TNTs, the Pt-in/5% Fe2O3-TNTs show a slightly lower temperature for complete oxidation of ethylene. Furthermore, the Pt-in/5% Fe2O3-TNTs demonstrate pretty good long-term stability, with no decay of catalytic activity over 24 h successive ethylene oxidation test at 95 °C, as shown in Figure b.

Figure 6

(a) Conversion curves for ethylene oxidation over the different catalysts and (b) reaction tests at 95 °C for ethylene oxidation with time-on-stream over Pt-in/5% Fe2O3-TNTs. Reaction conditions: C2H4 = 2.0 vol %, O2 = 49.0 vol %, Ar balance, GHSV = 24,000 mL/(g h).

Our previous studies have found that the confinement within TNTs can tune the electronic states of the encapsulated active component and strengthen the electrons transferred from the metal (Pt, Pd, or Au) to the nanotube surface.[21−23] In the meantime, the electronic properties of the encapsulated active component can be further regulated when an electron-withdrawing group is introduced, thus resulting in an enhanced catalytic activity.[28,29] In the present work, Fe2O3 was used to further regulate the electron density distribution of the confined Pt catalyst. Due to Fe2O3 with the property of binding electrons (Fe3+ + e → Fe2+), electrons are transferred from the interior to the exterior surface of TNTs. As a result, the entrapped Pt particles are stabilized at higher oxidation states and thus possess a great quantity of active oxygen species, which would make it much easier to adsorb reactant molecules and reduce their free energy, favoring the C2H4 oxidation reaction. On the contrary, after partial reduction of Fe2O3 to Fe3O4, the obtained Pt-in/5% Fe3O4-TNT catalyst exhibits the worse ethylene oxidation performance (see Figure S5). In order to characterize the presence of Fe3O4, the reduced FeO was obtained by a further reduction of pure Fe2O3 at 250 °C for 4 h under 8% H2–Ar atmosphere, and the corresponding XRD patterns were recorded. After the pure Fe2O3 was annealed under 8% H2–Ar atmosphere, the crystal phase structure of Fe3O4 (JCPDS no. 65-3107) can be identified in the XRD patterns, as shown in Figure S6. This further proves that heat treatment under 8% H2–Ar atmosphere at 250 °C for 4 h can partially reduce Fe2O3 to Fe3O4. It is inferred that Fe3O4 is an electron-donating component, which would suppress the electron migration from the inside to the outside surface of TNTs and hence exhibit a reduced oxidation activity. A higher percentage for ionic Ptδ+ species in the modified confined catalyst has been verified from the XPS analysis (see Figure and Table ). The Pt particles with higher oxidation states are believed to be promising for catalytic combustion.[10,28] As one-dimensional nanomaterials, TiO2 nanotubes have good electron transport characteristics. Owing to the unique structure of the catalyst with Pt nanoparticles inside and Fe2O3 on the outside surface of TNTs, the transformation of Fe3+ → Fe2+ could accelerate the electron transferred from Pt to Fe species via TiO2 nanotubes, facilitating the formation and migration of active oxygen species in the catalyst. Therefore, the catalytic activity for ethylene oxidation over the Pt-in/Fe2O3-TNTs is significantly improved by Fe2O3 modification.

Conclusions

In summary, Fe2O3-decorated TiO2 nanotube-confined Pt nanoparticles (Pt-in/Fe2O3-TNTs) was successfully synthesized by vacuum-assisted impregnation and deposition–precipitation methods. The Pt-in/Fe2O3-TNTs exhibit a significantly enhanced catalytic performance and a lower temperature for complete oxidation of ethylene than the Pt-in/TNTs. The improved catalytic oxidation performance can be ascribed to the modulation of electronic states for the encapsulated Pt particles by Fe2O3 modification on the exterior surface of TNTs. This study provides a prospect for VOC purification by Fe2O3-modified TiO2 nanotube-confined catalysts toward low-temperature applications.

Experimental Section

Catalyst Preparation

TNTs were synthesized via a hydrothermal method according to our previous report.[33] In a typical procedure, NaOH solution (10 M, 360 mL) and P25 powder (6 g) were added to a 500 mL Teflon-lined flask and refluxed at 110 °C under atmospheric pressure, keeping magnetic stirring for 48 h. The hydrothermally treated powders were collected and washed thoroughly with deionized water. The obtained precipitate was immersed into HCL solution (0.1 M) for 5 h and rinsed with deionized water until pH ∼ 7. The product was dried at 60 °C overnight and calcined at 350 °C for 2 h.

The Pt-in/TNT catalyst was prepared via a vacuum assisted incipient-wetness impregnation method. Briefly, 5.4 mL of H2PtCl6·6H2O ethanol solution (1.0 g/100 mL) was added dropwise into a round-bottom flask containing 1.0 g TNT powder to achieve nominal Pt loading of 2 wt %. The resulting slurry was vacuumed at room temperature until the excess solvent was evaporated completely. Finally, the dried powder was reduced at 250 °C for 4 h under 8% H2–Ar atmosphere.

The Pt-in/Fe2O3-TNT catalyst was prepared by a deposition–precipitation method. Typically, 0.5 g of Pt-in/TNT powders were dispersed in 40 mL of aqueous solution containing a calculated amount of Fe(NO3)3·9H2O. The pH was adjusted to ∼8 with 0.1 M Na2CO3 aqueous solution. The mixture was then aged at 60 °C for 3 h under stirring. After filtration and being washed with deionized water, the powder was dried at 60 °C and subsequently calcined at 250 °C for 4 h under air atmosphere. The nominal loading of Fe2O3 on TNTs was 1 and 5 wt % and labeled as Pt-in/1% Fe2O3-TNTs and Pt-in/5% Fe2O3-TNTs, respectively.

Catalyst Characterization

Surface area and porosity of the samples were analyzed by N2 adsorption–desorption using a SI-MP-10 automated system at liquid nitrogen temperature after the samples were degassed in a vacuum at 180 °C for 5 h. Powder XRD patterns were recorded using a PANalytical X’Pert Pro diffractometer with a Cu Kα radiation source operated at 40 kV and 40 mA. The morphologies of the samples were observed by HRTEM (FEI Tecnai G20). XPS was performed by a Thermo Fisher Scientific ESCALAB 250XI spectrometer with a monochromatized Al Kα source (1486.6 eV). The C 1s line was taken as an internal standard at 284.8 eV. The actual loading contents of Pt and Fe2O3 in the samples were determined by ICP-AES on a Plasma-Spec-I spectrometer.

Catalytic Activity Test

The catalytic oxidation performance for ethylene was tested in a continuous flow fixed-bed quartz reactor under atmospheric pressure. 100 mg samples were placed between two layers of quartz wool inside a quartz tube (i.d. = 6 mm) and heated at a rate of 1 °C/min and kept for 0.5 h under target temperature. The typical feed gas composition was 2% ethylene/49% O2 balanced with Ar with a total flow rate of 40 mL/min. Effluents (ethylene, CO, and CO2) from the reactor were analyzed by an Agilent 7890A gas chromatograph equipped with thermal conductivity and flame ionization detectors.