Formaldehyde (HCHO) emitted from newly installed building materials, plastic cements and paintings is one of the main air pollutants in the indoor environment, which seriously affects people’s health even at concentration lower than a few ppm ^[1-5]. As one of the most toxic air pollutants, HCHO has a negative influence on genetic material, respiratory tracts and cutaneous covering and also strongly affects the central nervous system ^[6]. Therefore, in order to improve indoor air quality and reduce public health risk, the abatement of HCHO at room temperature is a problem which should be urgently solved.

Over the past decades, many efforts have been made to investigate the removal of HCHO by environmental scientists ^[1]. Technologies that have been studied for the removal of HCHO are adsorption, photo-catalysis, plasma technology, and catalytic oxidation ^{[3, 7-11]}. Adsorption is a useful method for indoor air HCHO removal. Some absorbents can successfully reduce indoor HCHO concentrations to very low levels ^[12-14];however, their effectiveness is limited by adsorption capacity and they produce secondary pollution during regeneration. Photo-catalysis has received wide attention over the past decades and has been widely applied in energy production and environmental protection ^[15-18]. However, the photo- catalytic process requires the assistance of light energy, which may not available for the elimination of an indoor air pollutant and its wide application is limited. Catalytic oxidation can selectively decompose the toxic HCHO into harmless CO₂ and H₂Oat ambient temperature ^{[2, 19-21]}. Thus, the catalytic oxidation is regarded as the most promising method for indoor HCHO removal ^[1].

There are two major classes of catalysts for the complete oxidation of indoor HCHO: supported or unsupported transition metal oxides ^{[7, 8, 22, 23]} and supported noble metal catalysts ^{[3, 9-11]}. Of the former catalysts, the most efficient metal oxides that have been widely studied are MnO _x ^{[7, 24]}, CeO₂ ^[25], Co₃O₄ ^[26], and MnCeO _x ^[7]. Although they exhibit high activity at the beginning of the tests, these catalysts are easily deactivated in the presence of H₂O(H₂Ocan strongly adsorbed on the surface of the active sites, especially at low temperature), which greatly influences its practical application ^[7]. For most supported noble metal catalysts, their activity and water- tolerance are superior for volatile organic compounds (VOCs) elimination. Supported catalysts, such as Pt/TiO₂ ^{[3, 27]}, Pt/Al₂O₃ ^[28], Pd/TiO₂ ^[29], and Au/CeO₂ ^[6], have been widely studied and demonstrated high activity for HCHO oxidation. However, due to limited sources and the high price of noble metals, it is currently an important task and hot research topic to design and prepare catalysts with smaller noble metal particle size to reduce their loading.

For noble metal-based catalysts, it is commonly accepted that their activity is closely related to their particle size ^[30-33]. When studying the water gas shift reaction on Pd/CeO₂ and Pt/CeO₂, Gorte and coworkers ^[30] observed a linear relationship between activity and metal surface area. For methanol synthesis from H₂/CO on Pd/CeO₂, Shen and coworkers ^[31] also found that there was a close correlation between activity loss and the growth of the Pd particle size. In our previous studies, we also found that on Pd supported on various supports for CO and CH₄ deep oxidation that the activity is closed related to the Pd crystallite size and its metal surface area ^[34-36]. The same effect was also observed for Ni supported on different supports for CH₄ reforming for hydrogen production ^[37-39]. Therefore, the preparation of ultra- small noble metal nanoparticles of Pd, Pt, and Au has recently received a great deal of interest in the field of catalysis due to their size dependent catalytic activity ^[40-42]. Meanwhile, many previous results have demonstrated that the addition of certain amounts of rare earth oxides can improve the dispersion of supported noble metal catalysts and give much smaller particle size of the active component compared with the non-modified ones ^[43-46]. Therefore, the metal-support interaction of noble metal supported catalysts can be enhanced by rare earth modification ^{[44, 46-48]}.

To improve the utilization efficiency of noble metal-based catalysts, in the present study, anatase TiO₂ was modified by La first and then Pt was supported on it. This catalyst was applied for indoor low concentration HCHO oxidation at room temperature. To the best of our knowledge, there has been no report on the application of rare earth-modified Pt/TiO₂ for removing the low concentration HCHO at room temperature. The results demonstrated that La-doped TiO₂ remarkably reduced Pt particle size to below 2 nm. The La-modified TiO₂ was wash-coated on a cordierite monolith and then a very low amount of Pt (0.01 wt%) was supported on it. This catalyst was evaluated in a simulated indoor HCHO elimination environment, and it displayed high purifying efficiency and superior stability.

First, anatase TiO₂ was modified by various amounts of La (0, 1, 5, and 7 wt% La₂O₃) using a wet impregnation method. Typically, 0.4110 g La(NO₃)₃ was dissolved in an appropriate amount of deionized water. Then, 5 g TiO₂ (anatase, Aladdin) was added into the La(NO₃)₃ aqueous solution under constant stirring. After stirring for 2 h, excess water was removed in a rotary evaporator at 80 ℃ under vacuum. The samples were dried at 110 ℃ for 12 h and then calcined at 400 ℃ for 4 h. This catalyst was designed as 3 wt%La-TiO₂. The other amounts of La-modified TiO₂ was denoted as x wt%/La-TiO₂.

Second, Pt nanoparticles were supported on La-modified TiO₂ also by a wet impregnation method similar to that used for La-modified TiO₂. The amount of Pt was fixed at 0.5 wt% of the final catalyst, and Pt(NH₃)₄(NO₃)₂ was the Pt source. In detail, the 0.5 wt% Pt/ x wt% La-TiO₂ catalysts were prepared with an aqueous solution of Pt(NH₃)₄(NO₃)₂ instead of La(NO₃)₃ solution. An amount of the prepared x wt%La-TiO₂ support was measured and added into the Pt(NH₃)₄(NO₃)₂ solution. The mixture was constantly stirred overnight at 30 ℃ before removing excess water at 80 ℃, and dried at 110 ℃ for 12 h and calcined at 400 ℃ in an air atmosphere for 4 h. The sample was reduced by 10% H₂/N₂ at 400 ℃ for 4 h to obtain Pt supported on La-modified TiO₂, and designated as Pt/ x wt% La-TiO₂.

For commercial application, 3 wt%La-TiO₂ was wash-coated on cordierite monoliths (256 cpsi, 15 cm (length) × 15 cm (width) × 1.2 cm (height)) first and then a calculated amount of Pt (0.01 wt% based on the weight of the monolith) was impregnated onto this modified monolith. Before impregnation, the 3 wt%La-TiO₂ powder, which was calcined at 400 ℃ for 4 h in air, was wash-coated on the cordierite monolith. Prior to wash-coating, the monoliths were calcined at 500 ℃ for 2 h (with a ramp rate of 5 ℃/min) to remove adsorbed water. Then, a solid phase of 5 wt% boehmite (Disperal D) was dissolved in a slurry mixture consisting of liquid H₂Owas first used for the impregnation of the calcined monoliths in order to enhance the attachment of the 3 wt%La-TiO₂ powder. The alumina- coated monolith was previously calcined at 300 ℃ for 1 h and then 500 ℃ for 1 h (heating rate 2 ℃/min). Finally, the alumina-coated monoliths were immersed into a mixture of 3 wt%La-TiO₂ slurry (liquid to solid weight ratio = 80/20). The above procedure was repeated several times until the monoliths were coated with the desired amount of wash-coat. The La-TiO₂ coated monoliths were further calcined at 500 ℃ for 2 h at a heating rate of 5 ℃/min. Afterwards, La-TiO₂ coated monoliths was impregnated by 0.01 wt% Pt by conventional impregnation, where the amount of Pt was just 0.01% of the mass weight of the monolithic catalyst. Finally, the Pt/La-TiO₂ coated monoliths were calcined at 400 ℃ for 4 h, and then further reduced by 10% H₂/N₂ at 400 ℃ for 4 h.

N₂ adsorption-desorption isotherms were measured using a Micromeritics ASAP 2020 system. The specific surface area (S_BET) was estimated using the BET method. The pore size distribution was calculated using the Barrett-Joyner-Halenda (BJH) method, which calculated the pore size based on the adsorption layer thickness in relation to the pressure and the average radius of the meniscus obtained using the Kelvin method. Each sample was analyzed after degassing in vacuum at 300 ℃ for 4 h.

The powder X-ray diffraction (XRD) patterns were obtained on a Bruker AXS D8 Focus diffractometer operating at 40 kV and 40 mA with a Cu target and K_α X-ray irradiation. Scans were collected over a range of 2 θ from 10° to 90° with a step of 0.03°/s. The calculated experimental error for 2 θ measurement of the peaks was ± 0.01°.

CO temperature-programmed desorption (CO-TPD) was measured on an Auto-Chem 2920 (Micromeritics Co.) to analyze the Pt dispersion of the catalyst. The sample (typically 50 mg) was reduced first in a flow of 10% H₂/Ar gas mixture (30 mL/min) at 300 ℃ for 30 min, and then purged by a ultra-high purity He flow (30 mL/min) at room temperature for 30 min prior to CO adsorption in a 4% CO/He flow (20 mL/min). Afterwards, the sample was purged again in He flow (30 mL/min) for 30 min to remove physically adsorbed CO. After getting a stable baseline, the temperature was increased to 500 ℃ at a rate of 10 ℃/min. A CO desorption peak was generally detected for each sample at ~150 ℃, and the relative CO desorption amount was monitored by a thermal conductivity detector (TCD).

X-ray photoelectron spectroscopy (XPS) experiments were carried out on an RBD upgraded PHI-5000C ESCA system using a single Mg- K_α X-ray source operating at 250 W and 14 kV. The spectra were obtained at ambient temperature. The binding energies were calibrated using the C 1 s peak of graphite at 284.6 eV as reference.

The transmission electron microscope (TEM) and high angle dark field scanning transmission electron microscope (HAADFSTEM) images were taken on a Tecnai™ F30 field emission transmission electron microscope equipped with an energy dispersive spectroscopy (EDS) detector to determine the particle size of Pt.

The activity test for the catalytic oxidation of HCHO over the catalysts (100 mg) was performed in a fixed-bed quartz flow reactor (i.d. = 6 mm) in a chamber where the temperature was kept at 25 ℃, as shown in Fig. 1. Gaseous HCHO was generated by flowing air through the formaldehyde solution container in a water bath kept at 25 ℃. The feed gas composition was 0.5 ppm of HCHO, 21% O₂ and 50% RH balanced by nitrogen. The weight hourly space velocity (WHSV) was fixed at 60000 mL g^-1 h^-1. The inlet and outlet gases were monitored by INTERSCAN 4160 (detection range and limit are 0-19.99 and 0.01 ppm, respectively), which is a HCHO detector.

To test the performance of the monolithic catalyst, Pt supported cordierite monoliths were placed in an air purifying machine. Then, the machine equipped with monolithic catalyst was located in a 2 m³ sealed box to evaluate the purifying efficiency of HCHO in a real environment. The initial concentration of HCHO was fixed at about 0.5 ppm and monitored by a HCHO detector. The procedure for testing the performance of the monolithic catalysts is shown in Fig. 2.

Anatase TiO₂ was first doped with various amount of La₂O₃, and then 0.5 wt% Pt was supported on the La-modified TiO₂. The catalytic activity of Pt supported on La-modified TiO₂ powder catalyst for HCHO oxidation is shown in Figs. 3 and 4. The concentration of HCHO was fixed at 0.5 ppm, which is about 5 times higher than the allowed upper limit (0.08-0.1 ppm). This concentration is similar to the amount of HCHO emission in newly decorated buildings. There has been no report on HCHO removal using a concentration lower than 1 ppm. It is observed that the catalyst activity was promoted by the doped La as compared with the unmodified one (Fig. 4). The Pt/TiO₂ catalyst presented the lowest HCHO conversion at room temperature, just 88%. However, the conversions for the La-modified samples were all above 90%. The highest activity of nearly 100% at a GHSV of 60000 h^-1 at room temperature was obtained when the loading of La₂O₃ was 3.0 wt% (Fig. 4). Furthermore, the stability of Pt/3 wt%La-TiO₂ catalyst in the long term was tested at 25 ℃. As shown in Fig. 3(b), the sample exhibited excellent long term stability and catalytic activity, and the HCHO conversion was maintained at 100% for 40 h without any decrease. So, we concluded that anatase TiO₂ modified by La₂O₃ had a dramatic promotion effect for HCHO oxidation.

The textural properties of the Pt/3 wt%La-TiO₂ and Pt/TiO₂ catalysts were investigated together with pure TiO₂ and 3 wt%La-TiO₂ by N₂ adsorption-desorption measurements. The results are shown in Fig. 5 and the quantitative results listed in Table 1. As can be observed in Fig. 5, all the catalysts displayed type IV isotherm with type H3 hysteresis. The capillary condensation step observed in the isotherms is the characteristic of the presence of mesopores. Moreover, pure TiO₂ showed an obviously larger N₂ adsorption amount than the others. This was in line with the surface area and pore volume results in Table 1. Among all the supports, pure TiO₂ possesses the highest surface area (67 m²/g) and pore volume, while the surface area and pore volume showed a slight decrease after modification by La₂O₃ and doping Pt. The pore size distribution of the catalysts can be seen from the inset image in Fig. 5. All the samples showed two similar groups of pore diameter distribution: one group at 20 nm and the other at 1.6 nm. The former was assigned to the secondary aggregation of the particles, while the latter was assigned to particle micropores.

To identify the phase composition of the synthesized catalysts, all the samples were characterized by powder XRD. As the results show in Fig. 6, pure TiO₂ showed three diffraction peaks at 2 θ = 25.3° (101), 37.8° (004) and 48.0° (200), which are typical for the anatase TiO₂ phase. La species peaks were not observed with modified TiO₂, indicating that the La₂O₃ species were highly dispersed on the TiO₂ support. At the same time, the introduction of Pt did not induce the appearances of the peaks of polycrystalline Pt, which can be attributed to the low loading of Pt (0.5 wt%) and manifested that Pt nanoparticles with an ultra-small size at very high dispersion on the supports were obtained, which was further confirmed by the HRTEM and HAADF-STEM images and CO adsorption-desorption studies described in the following.

To better understand how the La₂O₃ modification affected Pt dispersion and particle size, CO-TPD, HRTEM and HAADFSTEM analysis of the catalysts were performed. Pt dispersion was determined by CO chemisorption. The results are listed in Table 1. Without La addition, the Pt dispersion on Pt/TiO₂ sample was 51%. La₂O₃ addition to Pt/TiO₂ remarkably increased the Pt dispersion to 66%. Clearly, La addition induced a more dispersed Pt species on Pt/ 3 wt%La-TiO₂ catalyst that exposed more Pt sites for HCHO oxidation, thus significantly promoting the activity of the catalyst.

Pt grain sizes were calculated based on the TEM and HAADF-STEM images shown in Figs. 7 and 8. The quantitative results are listed in Table 1. Fig. 7(a) and (b) showed that ultra- small Pt particles were well dispersed on the surface of the supports. The calculated Pt particle size was around 2.2 nm. Interestingly, the La₂O₃-modified samples showed a smaller Pt size of around 1.7 nm as shown in Fig. 7(c) and (d). The results were further confirmed by HAADF-STEM, as shown in Fig. 8. The images and results were in accordance with the Pt dispersion determined by CO chemisorption. All the characterization results demonstrated that the addition of La₂O₃ gave a dramatic improvement of Pt dispersion on TiO₂ support, leading to a smaller Pt particle size, more active sites exposed and superior activity for indoor low concentration HCHO oxidation. The turnover frequencies (TOFs) of the La-doped and undoped catalysts were 0.0776 and 0.0819 h^-1, respectively, which were very similar. This result further verified that the addition of La can effectively increase the dispersion of Pt and decrease the particle size of Pt.

XPS measurements were carried out to investigate the electronic state of the Pt species to reveal the effect of La₂O₃ on the Pt/TiO₂ catalyst. The results are shown in Fig. 9. Table 2 summarizes the binding energy (BE) values of Pt 4 f , Ti 2p and O 1 s and the atomic ratio of adsorbed O to lattice O calculated from the XPS results of the modified and unmodified catalysts. It is well known that the Pt 4 f_7/2 BE of Pt⁰, Pt²⁺ and Pt⁴⁺ are 71.1, 72.4 and 74.2 eV ^{[10, 11]}, respectively. For the Pt/ 3 wt%La-TiO₂ catalyst, the peaks of Pt 4 f were centered at 71.2 and 74.5 eV, so Pt⁰ and Pt⁴⁺ species simultaneously existed on this catalyst (Fig. 9(a)). It can be observed that the peaks of Pt 4 f of unmodified Pt/TiO₂ were very similar to those of La-modified Pt/TiO₂. From Fig. 9(b), we can also see that the peak of Ti 2p centered very closely at 457.8 eV. Since the La-modified Pt/TiO₂ did not affect the electronic state of Pt and Ti, the Pt dispersion and Pt particle size would be the most important factors for the catalytic performance in HCHO oxidation.

The O 1 s spectra of modified and unmodified Pt/TiO₂ are shown in Fig. 9(c). The O 1 s spectra exhibited two peaks: a main peak at 529 eV, and a shoulder peak at 531.0 eV. These results indicated two types of oxygen species existed on the catalysts. The former can be assigned to lattice oxygen (O_lat), and the latter to chemisorbed oxygen (O_ads). The O_ads/O_lat ratio was calculated from the peak areas. The results are shown in Table 2. It was found that O_ads/O_lat was very different between modified and unmodified Pt/TiO₂. The O_ads/O_lat ratio of Pt/La-TiO₂ was higher than that of Pt/TiO₂. Thus the La-doped TiO₂ can create more oxygen vacancies than the unmodified one, and the metal-support interaction would be enhanced due to the generation and activation of more chemisorbed oxygen species.

For commercial application, 3 wt%La-TiO₂ was wash-coated on cordierite monoliths and then 0.01 wt% Pt was loaded on the monoliths by a conventional impregnation method. The monolithic catalyst was tested in a 2 m³ sealed box to simulate the air environment for HCHO decomposition. The results are shown in Fig. 10. The initial HCHO concentration was 0.55 ppm, which is close to the real high concentration in an indoor environment. For the fresh catalysts, the HCHO concentration was decreased to 0.08 ppm within 5 min. Surprisingly, the catalyst maintain high activity even after 3 months, and the HCHO concentration decreased to 0.08 ppm in less than 10 min with the help of this catalyst. The catalytic performance of the monolithic catalyst showed that the La₂O₃ modified Pt/TiO₂ catalyst possessed high activity and stability in practical application. It would be a superior catalyst in the oxidation of indoor low concentration HCHO at ambient temperature.

This work demonstrated that La-modified TiO₂ had a dramatic promotion effect on a Pt/TiO₂ catalyst for ambient HCHO oxidation. The grain size of the Pt nanoparticles was reduced to below 2 nm after the TiO₂ support was modified by La₂O₃, which led to a higher Pt dispersion, more exposed active sites and thus superior activity for indoor low concentration HCHO oxidation. The XPS results showed that the amount of chemisorbed oxygen was more than on the unmodified support, so an enhanced metal-support interaction would be another reason for its higher activity. The La-modified TiO₂ was wash-coated on a cordierite monolith and then a very low amount of Pt (0.01 wt% of the total catalyst) was supported on it. This catalyst was evaluated in a simulated indoor HCHO elimination environment, and it displayed superior purifying efficiency and stability. Therefore, it is potentially a commercial catalyst for indoor HCHO elimination