Formaldehyde (HCHO) is a colorless gas with a strong irritating smell at atmospheric pressure. Outdoor HCHO mainly comes from the production of materials such as paints, textiles, printing materials, pesticides, and adhesives, and from motor vehicle exhausts. Indoor HCHO mainly comes from decorating materials, plywood, fiberboard, particleboard, and other artificial boards ^[1]. HCHO has serious adverse effects on human health and causes conditions such as edema, eye irritation, headaches, allergic dermatitis, and dark sports. Inhalation of HCHO at high concentrations can induce bronchial asthma, and HCHO can combine with protein amino groups to cause cell mutation. The damage caused by HCHO to human health is closely related to its concentration in air and contact time. It has been classified as carcinogenic and teratogenic by the World Health Organization ^{[2, 3]}. HCHO is also a volatile organic compound (VOC) and has strong photochemical activity, e.g., it can react photochemically with nitrogen oxides (NO_x) ^{[4, 5, 6, 7]}. HCHO removal is therefore necessary to protect human health and the atmospheric environment.

The main techniques used in the elimination of VOCs are adsorption, and photocatalytic and catalytic oxidation methods ^{[8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21]}. Adsorption usually uses activated carbon or molecular sieves as adsorbents for HCHO removal ^{[22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33]}. The use of this method is restricted because of the limitations of adsorption capacity and adsorbent regeneration. Photocatalytic methods often use TiO₂-based catalysts to remove HCHO ^{[34, 35, 36, 37, 38, 39, 40, 41]}. In actual applications, wall paints containing modified TiO₂ catalysts are used. However, under light, such paints can produce secondary pollution of toxicity similar to that of HCHO. Catalytic oxidation is a promising technique, and has advantages such as high removal efficiency, low light-off temperature, wide application scope, simple equipment, and no secondary pollution. HCHO can be directly converted to CO₂ and H₂O ^[42].

The development of catalytic oxidation techniques is important. The catalytic materials for HCHO oxidation are mainly divided into noble-metal and transition-metal oxide catalyst systems. In this paper, we review progress in research on these two systems in detail, and future directions and potential hotspots in research on catalysts for HCHO oxidation are discussed.

Noble-metal catalyst attract much attention because of their excellent low-temperature oxidation activities. They are loaded on supports because precious metals themselves are easily volatilized, oxidized, and sintered. The loading of precious metals on supports enables HCHO conversion at low temperatures. The specific catalytic properties are related to factors such as precious-metal and support types, and structure. The precious-metal catalysts currently used for HCHO oxidation mainly contain Pt, Pd, Au, and Ag as the active components ^[43]. Other precious metals are not suitable for catalytic combustion because of their high volatilities and ease of oxidation at high temperatures. The supports for precious-metal catalysts for HCHO oxidation can be divided into three types. The first type is materials with no oxidation activities and large specific surface areas, such as SiO₂, Al₂O₃, TiO₂, and molecular sieves; these are common catalyst supports and are commercially available. The second type is single or mixed metal oxides without special morphologies, with high-temperature oxidation activities, but low specific surface areas; examples are bulk CeO₂ and MnO₂; these are traditional metal oxide supports. The third type is metal oxides with special morphologies such as nanorods and nanospheres, and nanoporous materials; these have better catalytic activities than the corresponding traditional bulk metal oxides. These materials, e.g., MnO₂ nanorods or nanospheres, CeO₂ nanospheres, mesoporous Co₃O₄, and macroporous CeO₂, are metal oxide supports with special morphologies. Table 1 gives an overview of the catalytic activities in HCHO oxidation of various supported noble-metal catalysts.

Table 1 Overview of catalytic activities in HCHO oxidation of supported noble-metal catalysts. Catalysts Reaction conditions T50 (oC) Ref. Common supports Pt/TiO2 Rh/TiO2 Pd/TiO2 Au/TiO2 100 ppm HCHO, 20 vol% O2, 50000 h-1 SV R.T. 50 70 90 [44,45] Na-Pt/TiO2 600 ppm HCHO, 20 vol% O2, 300000 h-1 SV, 50% relative humidity R.T. [46] Na-Pt/TiO2 105 ppm HCHO and 315 ppm toluene R.T. [47] Pt/TiO2(C) 36 ppm HCHO, 21 vol% O2, total flow 500 mL/min R.T. [48] Pt/f-SiO2 Pt/SBA-15 Pt/p-SiO2 300 ppm HCHO, 20 vol% O2, 30000 mL/(g·h) SV R.T. 40 90 [49] Pt/TiO2 Pd/TiO2 Rh/TiO2 Pt/SiO2 100 ppm HCHO, 22 vol% O2, 300000 mL/(g·h) SV R.T. 80 90 60 [50] Pt/carbon 100-300 ppm HCHO, 22 vol% O2, 1120 h-1 SV <100 [51] Pt/TiO2 10 ppm HCHO, 80000 h-1 SV, 50% relative humidity R.T. [52] Pd/TiO2 10 ppm HCHO, 120000 h-1 SV, 50% relative humidity R.T. [53] Pd/Bata Pd/USY Pd/ZSM-5 Pd/HM10 Pd/Zeo-13X Pd/Al2O3 40 ppm HCHO, 20 vol% O2, 50000 h-1 SV, 3% H2O <40 <40 70 >140 <40 <40 [54] PdMn/Al2O3 0.5% HCHO, 0.2% CH3OH, 0.7% H2O, 75.6% N2, 23% O2 60 [55] Ag/SBA-15 1000 ppm HCHO, 15 vol% O2, 15000 mL/(g·h) SV 50 [56] Ag/Al2O3 Ag/SiO2 1.2% HCHO, 14.8% O2, 1000 or 7000 h-1 SV — [57] Ru/Al2O3 Ru/zeolite Ru/TiO2 900 ppm HCHO, 160 ppm CH3OH, 18% H2O, 82% air, 20000 h-1 SV 198 210 212 [58] Traditional metal oxide supports Pt/CeO2 Pd/CeO2 Rh/CeO2 Ru/CeO2 Rh/ZrO2 Ir/CeO2 900 ppm HCHO, 160 ppm CH3OH, 18% H2O, 82% air, 20000 h-1 SV <150 <150 <150 <150 188 207 [58] Ag/CeO2 0.42% HCHO, 0.074% CH3OH, 19.9% H2O, 62.7% N2, 16.9% O2, 21000 h-1 SV <150 [61] Pt/MnOx-CeO2 580 ppm HCHO, 20 vol% O2, 30000 mL/(g·h) SV R.T. [62] Ag/MnOx-CeO2 Ag/CeO2 Ag/MnOx 580 ppm HCHO, 18 vol% O2, 30000 mL/(g·h) SV 70 90 115 [63] Pt/CeZrO2 100 ppm HCHO, 22 vol% O2, 300000 mL/(g·h) SV >120 [50] Pd/Mn-CeO2 40 ppm HCHO, 50000 mL/(g·h) SV, 3% H2O 110 [54] Au/CeO2 0.06% HCHO, 32000 mL/(g·h) SV 90 [64] Au/CeO2 500 ppm HCHO, 20 vol% O2, 35400 h-1 SV <40 [65] Au/Fe-O 6.25 mg/m3 HCHO, 54000 mL/(g·h) SV 40 [66] Pt/Fe2O3 100-500 ppm HCHO, 60000 mL/(g·h) SV R.T. [67] Pt/K-OMS Ag/K-OMS 460 ppm HCHO, 21 vol% O2, 30000 mL/(g·h) SV >200 >200 [68] Metal oxide supports with special morphologies Pt/nest-like MnO2 460 ppm HCHO, 20000 mL/(g·h) SV R.T. [69] Ag-HMO 400 ppm HCHO, 10 vol% O2, 30000 mL/(g·h) SV 80 [70, 71] 3DOM Au/CeO2 0.06% HCHO, 66000 mL/(g·h) SV 70 [72, 73] 3DOM Au/CeO2-Co3O4 3DOM Au/CeO2 3DOM Au/Co3O4 0.06% HCHO (8ppm), 15000 mL/(g·h) SV R.T. 42 30 [74] 2D Au/Co3O4-CeO2 2D Au/Co3O4 2D Au/CeO2 200 ppm HCHO, 22 vol% O2, 55000 h-1 SV R.T. R.T. — [75] Mesoporous Au/ZrO2 90 mg/m3 HCHO, 52000 mL/(g·h) SV 120 [77] 3D Ag/Co3O4 3D K-Ag/Co3O4 100 ppm HCHO, 20 vol% O2, 30000 h-1 SV <80 R.T. [78] Ag/CeO2 nanospheres Ag/CeO2 bulk particles 810 ppm HCHO, 20 vol% O2, 84000 h-1 SV 90 140 [79] Pt-Ce/OMS-2 Pt/OMS-2 500 ppm HCHO, 10 vol% O2, 30000 mL/(g·h) SV 120 100 [80]

Table 1 Overview of catalytic activities in HCHO oxidation of supported noble-metal catalysts.

Common supports used in HCHO oxidation are materials such as SiO₂, Al₂O₃, TiO₂, and molecular sieves. Noble-metal catalysts are often prepared by loading Pt, Pd, Rh, Ru, Au, and Ag on common carriers with large surface areas.

Zhang et al. ^{[44, 45]} reported TiO₂-supported noble-metal (Pt, Rh, Pd, and Au) catalysts. The order of the HCHO oxidation activities over TiO₂-supported noble-metal catalyst was Pt/TiO₂ > Rh/TiO₂ > Pd/TiO₂ > Au/TiO₂. Complete HCHO conversion to CO₂ and H₂O was achieved over 1% Pt/TiO₂ at 20 °C at a gas hourly space velocity (GHSV) of 50000 h^-1 (Fig. 1). High Pt dispersion is an important reason for the high activities. Based on the results achieved using a 1% Pt/TiO₂ catalyst, this research group also prepared 2% Na-1% Pt/TiO₂ using Na⁺ ions as auxiliaries ^[46]. This catalyst has the best catalytic activity and stability reported to date for HCHO oxidation, and complete conversion is achieved at room temperature for 600 ppm HCHO and a GHSV of 300000 h^-1 (Fig. 2). The activity is high because Na⁺ ions generate surface hydroxyl groups for oxidation, and the presence of surface hydroxyl groups changes the reaction mechanism. Nie et al. ^[47] reported Pt/TiO₂ catalysts with various Pt loadings (0.05-2 wt%). They tested their catalytic performances in HCHO oxidation in the presence of toluene. HCHO was selectively oxidized to CO₂ and H₂O by the Pt/TiO₂ catalysts, and toluene did not affect the catalytic reaction. Pt/TiO₂ catalysts prepared with the assistance of NaOH showed higher HCHO oxidation activities than those without NaOH because of the presence of surface hydroxyl groups, high adsorption capacity, and large mesopores and macropores, which facilitate diffusion and transport of reactants and products. Kim et al. ^[48] reported HCHO oxidation at room temperature on Pt/TiO₂ catalysts. The results show that the catalytic activity depends on the type of TiO₂ carrier, metal valence state of the catalyst, and HCHO feed rate. CO is the main intermediate species in HCHO oxidation. An et al. ^[49] prepared Pt/SiO₂ and Pt/SBA-15 catalysts; these completely converted HCHO at room temperature and 50 °C, respectively. The use of SiO₂ as a carrier may affect the particle size and chemical valence of Pt species, and this can be favorable to redox reactions of Pt/SiO₂. Peng et al. ^[50] also prepared supported noble-metal (Pt, Pd, Rh) catalysts. The activity order over TiO₂-based catalysts was Pt/TiO₂ > Pd/TiO₂ > Rh/TiO₂. The activity order for various supports loaded with Pt was TiO₂ > SiO₂ > CeZrO₂. Chuang et al. ^[51] proposed a reaction mechanism for HCHO oxidation over a Pt catalyst supported on hydrophobic fluorinated carbon. HCHO conversion (T₅₀) over the Pt/carbon catalyst was achieved below 100 °C. Huang et al. ^[52] prepared Pt/TiO₂ catalysts and examined the effects of metal oxidation state, carrier, and Pt particle size on the catalytic performance in HCHO oxidation. A catalyst consisting of 0.1 wt% Pt/TiO₂ had excellent activity in HCHO oxidation, because the reduction of Pt particles increased electron mobility and the number of reactive oxygen species. This group also prepared Pd/TiO₂ catalysts ^[53]. These catalysts gave better catalytic performances in HCHO oxidation, with turnover frequencies (TOFs) of 0.015 s^-1 at room temperature. Strong metal-support interactions, well-dispersed and negatively charged metallic Pd nanoparticles, and large amounts of chemically adsorbed oxygen are probably responsible for the high catalytic activities. In the reaction, Pd nanoparticles provide active sites for HCHO oxidation, and promote oxygen activation. Park et al. ^[54] prepared a series of Pd catalysts supported on Bata, USY, ZSM-5, HM10, and Zeo-13X molecular sieves. Pd/ZSM-5 Pd/Bata, Pd/USY, Pd/HM, and Pd/Zeo-13X all have good HCHO catalytic activities at low temperatures (T₅₀ < 40 °C). The catalytic performance in HCHO oxidation of the Pd/Bata catalyst was slightly better than those of the others. Formate and dioxymethylene are the main intermediates in the reaction, and these are the surface species involved in HCHO oxidation. dela O’Shea et al. ^[55] prepared a Pd-Mn/Al₂O₃ catalyst, and investigated the effect of feed concentration on its catalytic performance in HCHO/methanol oxidation. Its catalytic activity was low because of decomposition of PdO to Pd and MnO₂/Mn₂O₃ to Mn₃O₄. There were fewer oxygen species available for transfer to Pd⁰. In addition, after water vapor addition, catalyst passivation resulted in competition between water and VOC molecules for active sites. Qu et al. ^[56] investigated the oxidation of HCHO on a Ag/SBA-15 catalyst. The catalyst completely converted HCHO at 100 °C at a GHSV of 15000 mL/(g·h) (Table 1). This is attributed to higher Ag dispersion in the narrow pores of SBA-15 and high HCHO adsorption. Ag nanoparticles provide active sites for oxidation. In the absence of Ag, HCHO is only converted to formate and dioxymethylene. Mao et al. ^[57] prepared Ag/Al₂O₃ and Ag/SiO₂ catalysts, and clarified the mechanism of HCHO oxidation. The formation of CO₂ and CO is attributed to the decomposition of formate on the catalyst surface, and the formation of formate is attributed to the reaction of dioxymethylene.

Fig. 1. HCHO conversions on TiO2-supported 1 wt% noble-metal (Pt, Rh, Pd, and Au) catalysts at various temperatures [45]. Reaction conditions: HCHO 100 ppm, O2 20 vol%, He balance, total flow rate 50 mL/min, GHSV 50000 h-1

Fig. 2. HCHO conversions over Na-Pt/TiO2 catalysts [46]. Reaction conditions: HCHO 600 ppm, O2 20 vol%, relative humidity 50%, He balance, total flow rate 50 mL/min, and GHSV 120000 h-1 (inset: 2% Na-1% Pt/TiO2 stability test at 25 °C, GHSV 300000 h-1, other reaction conditions the same).

Noble metals loaded on common supports with large surface areas have an excellent catalytic activity. This is related to the specific surface areas of the supports, noble-metal dispersion, active sites, particle size, valence state, and adsorption ability. A large surface area increases exposure of a large number of active sites and the adsorption and diffusion of reactants and products, and can enhance the synergistic effect between the support and the active component. Table 1 shows that the catalytic activities in HCHO oxidation of various noble metals loaded on common supports decrease in the following order: Pt > Pd > Rh > Au > Ag. Pt/TiO₂ ^[45], and Pd/TiO₂ ^[53] can completely convert HCHO at room temperature. Na-Pt/TiO₂ is the best catalyst for HCHO removal ^[46]. Generally, use of a different support but the same active components affects the catalytic activity. For example, Imamura et al. ^[58] reported the oxidation of HCHO over supported noble-metal catalysts. HCHO conversions (T₅₀) using Ru/Al₂O₃, Ru/zeolite, and Ru/TiO₂ were achieved at 198, 210, and 212 °C, respectively, and conversion with Ru/CeO₂ was obtained below 150 °C. The HCHO oxidation activity of Ru/CeO₂ is better than those of Ru/Al₂O₃, Ru/zeolite, and Ru/TiO₂. For CeO₂ loaded with Ru, Pd, Rh, Pt, and Ir, the order of the catalytic activities (T₉₀) is Ru/CeO₂ > Pd/CeO₂ > Rh/CeO₂ > Ir/CeO₂ > Pt/CeO₂. The better HCHO oxidation activity of the Ru/CeO₂ catalyst indicates that other noble metals, as well as Pt and Pd, may have potential catalytic applications if an appropriate common support is used.

Catalysts that have the advantages of both precious-metal and metal oxide catalysts can be obtained by loading precious metals on metal oxide supports. Such catalysts have better low-temperature oxidation activities and thermostabilities, and enhanced interactions between the metal and the support. Sekizawl et al. ^[59] reported a Pd/SnO₂ catalyst. The results show that the catalyst has a lower specific surface area, but better hydrothermal stability than a Pd/Al₂O₃ catalyst. Minico et al. ^[60] reported that Au nanoparticles can produce synergies with CeO₂ and Fe₂O₃, reduce the combining capacities of oxygen, Ce, and Fe, strengthen the interactions between Au and the support, and promote reactions with surface oxygen species.

Traditional metal oxide supports for the catalytic oxidation of HCHO, such as CeO₂, Fe₂O₃, Co₃O₄, and MnO₂, or their composites, are usually prepared by precipitation, coprecipitation, or sol-gel methods. Imamura et al. ^{[58, 61]} reported that supported noble-metal (Ag, Ru, Pd, Rh, and Pt) catalysts all displayed activity in HCHO oxidation below 150 °C. Tang et al. ^[62] prepared Pt/MnO_x-CeO₂ catalysts and examined the effects of the two precursors on the HCHO oxidation performance. The results show that the activity and stability of the Pt-containing catalyst prepared from Pt(NH₃)₂(NO₂)₂ are better than those of the catalyst prepared from H₂PtCl₆. It can completely oxidize HCHO at room temperature because of effective activation of oxygen molecules on the MnO_x-CeO₂ carrier. A Ag/MnO_x-CeO₂ catalyst was also prepared ^[63]. Its HCHO oxidation activity (Table 1) is better than those of Ag/MnO_x and Ag/CeO₂. This is attributed to formation of a MnO_x-CeO₂ solid solution and Ag₂O oxygen species produced from a Mn⁴⁺/Mn³⁺ and Ce⁴⁺/Ce³⁺ redox cycle. Shen et al. ^[64] reported a series of Au/CeO₂ catalysts in which Au clusters are dispersed on the catalyst surfaces. When the Au content is 0.78%, the Au particle size is about 10 nm. The results show that large Au particles are not beneficial to catalytic oxidation. Highly dispersed Au crystallite clusters can provide more active sites for HCHO oxidation. Li et al. ^[65] also reported the catalytic oxidation of HCHO on Au/CeO₂. An increase in the specific surface area enhanced the catalytic ability. The catalyst has better oxidation activity for two reasons. One is the high valence of Au species on the CeO₂ surface, the other is formation of oxygen vacancies and a Au_xCe_1-xO_2-δ solid solution. Li et al. ^[66] prepared a catalyst with Au as the active component and iron oxide as the carrier. The HCHO oxidation activity of the 7.10% Au/Fe-O catalyst (T₅₀ = 40 °C, Table 1) was the best. This is attributed to the presence of Au^δ+ and active species that play an important role in the reaction. An et al. ^[67] reported that Pt/Fe₂O₃ catalysts after calcination at 200 and 300 °C have good activities and stabilities. HCHO can be completely converted at room temperature. This is attributed to interactions between Pt and Fe₂O₃ to form Pt-O-Fe bonds, which are favorable for oxidation. Tian et al. ^[68] prepared K-OMS-2 using a sol-gel method and loaded Pt and Ag on it using a conventional impregnation method. Unlike the cases for other reported catalysts such as Pt/MnO₂ ^[69] and Ag/MnO₂ ^[70], the addition of Pt or Ag reduced the catalytic activity. The order of the catalytic activities in HCHO oxidation was K-OMS-2 > Pt/K-OMS > Ag/K-OMS. HCHO conversion of 50% was achieved over K-OMS-2 at 180 °C.

Among the catalysts of this type listed in Table 1, supported Pt catalysts have excellent catalytic activities (T₅₀) at room temperature. However, some catalysts such as Au/CeO₂, Ag/MnO_x-CeO₂, and Ag/CeO₂ have better development potential, with HCHO conversions (T₅₀) achieved below 40 °C on Au/CeO₂ ^[65], at 70 °C on Ag/MnO_x-CeO₂, and at 90 °C on Ag/CeO₂ ^[63]. These results are attributed to stronger interactions between the noble metal and the support, and special crystalline structures with solid solutions. The choice of support is very important for the catalytic activity in HCHO oxidation. The same Au/CeO₂ catalysts ^{[64, 65]}, but with different specific surface areas, show different catalytic activities. Au/CeO₂ with a higher surface area clearly has better oxidation activity (T₅₀ < 40°C); the surface area is related to the preparation method. A range of supports are available for supported noble-metal catalysts. Many studies have confirmed that supported Pt catalysts are promising, but limited by the cost of Pt, therefore Pt may not be the best choice as the active component for HCHO removal. Some supported Au and Ag catalysts could be used as alternatives to Pt catalysts. The metal oxide supports can also be modified by using different preparation methods.

Catalysts supported on metal oxides with special morphologies, which are mainly prepared using hydrothermal and hard template methods, have higher catalytic activities than those supported on conventional bulk metal oxides prepared using precipitation methods. Catalysts with specific morphologies, based on noble metals supported on metal oxides with special morphologies, are better able to remove HCHO at low temperatures, even room temperature, than catalysts with traditional metal oxide supports. Many metal oxides with special morphologies such as tubes, lines, rods, sheets, flowers, spheres, cubes, and pores have been reported, but only metal oxide supports with morphologies involving rods, spheres, mesopores, and macropores have been used in catalytic oxidation of HCHO.

Yu et al. ^[69] reported a Pt/MnO₂ catalyst with good HCHO oxidation activity. Pt nanoparticles were evenly dispersed on the surfaces of MnO₂ nanospheres. The addition of Pt clearly reduced the reaction temperature. The Pt dispersion, grain size, and interactions between Pt and MnO₂ are the main reasons for the good catalytic activity in HCHO oxidation. Tang’s group ^[70] reported a Ag-hollandite manganese oxide (HMO) catalyst with single Ag atoms as active sites. The HMO nanorods consist of one-dimensional 0.47 × 0.47 nm² square tunnels, and the basic unit of the tunnel structure is built from eight (4 + 4) oxygen atoms to form a tetragonal prism. Each of the eight oxygen atoms in the tunnels has four sp³-hybridized orbitals, three of which bond to three Mn⁴⁺ cations and the other, occupied by lone-pair electrons, points to the central axis of the tunnel. Theoretically, linear tunnels of a specific size and electron-rich tunnel oxygen atoms can serve as a natural mold and an electron donor, respectively; this favors formation of stable single-atom Ag chains in the HMO tunnels. The single Ag atoms anchored at the openings of the Ag-HMO tunnel nanorods are active sites, and can easily activate lattice oxygen and molecular oxygen. The Ag-HMO structure is shown in Fig. 3. The activation of oxygen species results in excellent HCHO oxidation activity at low temperatures. The Ag-HMO catalyst has a high TOF, 0.054 s^-1, at 90 °C and HCHO conversion (T₅₀) is achieved at 80 °C (Table 1). This group also examined the atomic Ag centers ^[71]. Zhang et al. ^[72] prepared three-dimensional ordered macroporous (3DOM) Au/CeO₂ with controlled pore sizes. Fig. 4 shows scanning electron microscopy (SEM) and transmission electron microscopy (TEM) images, and HCHO conversions for 3DOM 1 wt% Au/CeO₂ prepared using CeO₂ of pore size 80 nm. The catalyst has interconnected networks of spherical voids. Au nanoparticles are uniformly dispersed on the surface. The catalytic activity of 3DOM 1 wt% Au/CeO₂ in HCHO oxidation is excellent and complete conversion of 0.06 vol% HCHO is achieved at 75 °C. This temperature is lower than that needed for complete conversion using non-porous Au/CeO₂. The better activity of 3DOM Au/CeO₂ in HCHO oxidation is the result of Au dispersion, a higher Au content, and the presence of Au³⁺. This group also reported the mechanism of HCHO oxidation on the 3DOM Au/CeO₂ catalyst ^[73]. HCHO molecules are adsorbed to form formate species on the catalyst surface. The adsorption and activation of oxygen species are related to the Au³⁺/Au⁰ and Ce⁴⁺/Ce³⁺ redox cycles. During HCHO oxidation, formate species are further oxidized to CO₂ and H₂O by active oxygen species. If the formate is not completely oxidized, carbonate or bicarbonate is produced on the 3DOM Au/CeO₂ surface, and may cause catalyst passivation. However, it is difficult to block the active sites on the catalyst surface because of the large and open pores. A 3DOM Au/CeO₂-Co₃O₄ catalyst was prepared by this group, based on 3DOM Au/CeO₂ ^[74]. HCHO can be completely converted at 39 °C. The interactions between CeO₂ and Co₃O₄ promote the migration of surface oxygen species and activation of Au species. Ma et al. ^[75] reported that two-dimensional (2D) ordered mesoporous Au/Co₃O₄-CeO₂ had good catalytic activity in HCHO oxidation, with 50% HCHO conversion at room temperature (Table 1). The catalytic activity decreases with increasing CeO₂ content. The (110) crystal planes of Co₃O₄ are the main active faces, and they can adsorb and activate HCHO species. Au loading facilitates the formation of surface oxygen species. This group also prepared a similar 3D Au/Co₃O₄ catalyst for the elimination of ethylene, and a better catalytic performance was obtained ^[76]. Zhang et al. ^[77] prepared mesoporous Au/ZrO₂ for HCHO oxidation. The Au is better dispersed, with more Au³⁺ ions and a stronger HCHO adsorption ability. The Au³⁺ ions are reduced to Au⁰ metal, and adsorbed HCHO molecules are quickly converted to formate species. A catalysts consisting of 0.85 wt% Au/ZrO₂ can completely convert HCHO at 180 °C because of adsorption of HCHO on Au species and of oxygen molecules on the support.

Fig. 3. TEM and high-resolution TEM images and corresponding structural models of Ag–HMO; TOFs for HCHO oxidation over Ag–HMO are taken from Ref. [70].

Fig. 4. SEM and TEM images of 3DOM Au ~1 wt%/CeO2; HCHO conversions of (a) 3DOM Au ~1 wt%/CeO2 (80 nm pore size), (b) 3DOM Au ~1 wt%/CeO2 (130 nm pore size), (c) 3DOM CeO2 (80 nm pore size), (d) 3DOM CeO2 (130 nm pore size) from Ref. [72].

Our group used KIT-6 mesoporous silica as a hard template to prepare 3D ordered Co₃O₄ ^[42]. Mesoporous metal oxides have potential as catalysts. Mesoporous Ag/Co₃O₄ and K-Ag/Co₃O₄ catalysts based on 3D Co₃O₄ were prepared ^[78]. Ag nanoparticles were uniformly dispersed and supported on a polycrystalline wall. The addition of K⁺ ions strengthens anionic lattice defects and interactions between Ag and the Co₃O₄ support, resulting in formation of more Co³⁺ ions and surface lattice oxygen species. The catalytic activity in HCHO oxidation of the K-Ag/Co₃O₄ sample is better than that of the Ag/Co₃O₄ catalyst because of the presence of surface hydroxyl species and exposure of Ag (111) active faces. HCHO conversion of 50% was achieved at room temperature using K-Ag/Co₃O₄ (Fig. 5). Our group reported catalysts with Ag nanoparticles loaded on CeO₂ nanospheres and bulk particles ^[79]. Fig. 6 shows TEM images of the Ag/CeO₂ nanospheres and the achieved HCHO conversions. The catalytic activity in HCHO oxidation of Ag/CeO₂ nanospheres is better than that of Ag/CeO₂ bulk particles. Ag/CeO₂ nanospheres can completely convert 810 ppm HCHO at 110 °C and a GHSV of 84000 h^-1; the HCHO conversion (T₅₀) at 90 °C is shown in Table 1. The reaction rate is almost 3.6 times that obtained with Ag/CeO₂ bulk particles. The average Ag/CeO₂ nanosphere size is between 80 and 100 nm, and it consists of small crystallites of size 2-5 nm. This special structure promotes good distribution of Ce and Ag. Oxygen is easily chemisorbed on the Ag/CeO₂ nanosphere surfaces, and interactions between Ag and CeO₂ may occur in the catalyst. Our group also reported that the catalytic activity of Pt-Ce/OMS-2 in HCHO oxidation ^[80] was lower than those over Ag-HMO ^[70], K-Ag/Co₃O₄ ^[78], and Ag-CeO₂ nanospheres ^[79].

Fig. 5. HCHO conversion over K-Ag/Co3O4 catalyst [78].

Fig. 6. TEM image of (a) CeO2-N nanosphere, (b) Ag/CeO2-N nanosphere, (c) Ag/CeO2-N nanosphere, (e) Ag/CeO2-N nanosphere and (d) the distribution of element maps of Ce (green) and Ag (yellow) on Ag/CeO2-N nanosphere and HCHO conversions over Ag/CeO2 nanospheres [79].

Table 1 shows that metal oxide supports with special structures have rarely been used for Pt loading, except Pt/nest-like MnO₂ ^[69] and Pt/OMS-2 ^[80]. This is because 2% Na-1% Pt/TiO₂ has become the benchmark for supported Pt catalysts, and is used in air purification. More attention has been paid to catalysts consisting of Ag and Au loaded on metal oxide supports with special morphologies. Such catalysts, e.g., mesoporous 3DOM Au/CeO₂-Co₃O₄ ^[74], 2D Au/Co₃O₄-CeO₂, 2D Au/Co₃O₄ ^[75], and 3D K-Ag/Co₃O₄ ^[78], all have excellent catalytic activities in HCHO oxidation at room temperature (T₅₀ = room temperature) and have good potential applications. The interactions between the noble metal and the support, dispersion, particle size, low-temperature reducibility, mobility and activation of active oxygen species, and redox cycles of high- and low-valence states of metal ions are responsible for the excellent catalytic activities in HCHO oxidation. Experimental results have shown that Au and Ag, which are cheaper than Pt and Pd, can provide sufficient active sites for HCHO oxidation (Table 1). Catalysts with Au and Ag on supports with special morphologies, especially porous structures, are promising for HCHO removal at room temperature and may become substitutes for Pt catalysts.

Precious-metal catalysts have excellent low-temperature catalytic activities in HCHO oxidation, but their industrial use is restricted because of their high costs. Precious-metal catalysts have poor thermal stabilities at 500-900 °C. These catalysts are easily sintered, resulting in decreased surface area, which decreases their catalytic activities. Transition-metal oxides are cheap and plentiful, and transition-metal oxide catalysts have potential applications in VOC removal. Transition-metal oxide catalysts with good thermal stabilities have therefore attracted much attention. Transition-metal oxide catalysts with good activities (T₁₀₀ ≤ 140 °C) in HCHO oxidation mainly contain Mn, Co, and Cr. Other metal elements such as Ce, Sn, Cu, and Zr can be doped into MnO_x and Co₃O₄ to prepare composite metal oxides if a single metal oxide does not give a good catalytic performance. In addition, a single rare-earth oxide, i.e., CeO₂, has been used in the catalytic oxidation of HCHO. Table 2 gives an overview of the catalytic activities of transition-metal oxides in HCHO oxidation.

Table 2 Overview of catalytic activities of transition-metal oxides in HCHO oxidation. Catalysts Reaction conditions T50 (oC) Ref. Single metal oxide catalysts MnO2, Ag2O, PdO, CoO, CuO, ZnO, Fe2O3, La2O3, V2O5, TiO2, CeO2 and Mn3O4 <750 ppm HCHO — [81, 82] Cryptomelane Birnessite Ramsdellite MnOOH nanorods 100 ppm HCHO, 20 vol% O2, 50000 h-1 SV 90 95 100 115 [83] Pyrolusite Cryptomelane Todorokite 400 ppm HCHO, 10 vol% O2, 18000 mL/(g·h) SV 150 110 140 [84] Cocoon-like MnO2 Urchin-like MnO2 Nest-like MnO2 460 ppm HCHO, 20000 mL/(g·h) SV 130 120 110 [69] Hollow KxMnO2 Honeycomb MnO2 100 ppm HCHO, 20 vol% O2, 50000 h-1 SV 50 70 [88] OMS-2 nanorods MnOx 0.01% HCHO, 20 vol% O2, 24000 mL/(g·h) SV 75 85 [89] OMS-2 500 ppm HCHO, 10 vol% O2, 30000 mL/(g·h) SV 110 [80] OMS-2 nanoparticles OMS-2 nanorods 460 ppm HCHO, 21 vol% O2, 20000 mL/(g·h) SV 90 >100 [90] Birnessite MnO2 460 ppm HCHO, 21 vol% O2, 30000 mL/(g·h) SV 80 [91] 3D-Cr2O3 500 ppm HCHO, 30000 mL/(g·h) SV 92 [92] 3DOM CeO2 0.06% HCHO, 66000 mL/(g·h) SV 250 [72] 3D-Co3O4 2D-Co3O4 Co3O4 nanopaticles 400 ppm HCHO, 20 vol% O2, 30000 mL/(g·h) SV <110 <130 <210 [42] CeO2 nanospheres CeO2 bulk particles 810 ppm HCHO, 20 vol% O2, 84000 h-1 SV 175 >250 [79] 3D-MnO2 α-MnO2 nanorods β-MnO2 nanorods 400 ppm HCHO, 20 vol% O2, 30000 mL/(g·h) SV <100 100 >140 [94] Composite metal oxide catalysts MnOx-CeO2 (Mn/Ce = 1) 580 ppm HCHO, 20 vol% O2, 30000 mL/(g·h) SV 580 ppm HCHO, 18 vol% O2, 21000 mL/(g·h) SV <90 [62] [95] CeMn10 CeMn30 CeMn50 CeMn80 MnOx CeO2 580 ppm HCHO, 20 vol% O2, 150 125 100 80 72 225 [96] MnOx-SnO2 400 ppm HCHO, 10 vol% O2, 30000 mL/(g·h) SV 150 [97] CuO/MnO2 0.18-1.3 ppm HCHO, 100000 h-1 SV <120 [98] Co(N)/Zr Co-CD0.1/Zr 150 ppm HCHO, 20 vol% O2, 30000 mL/(g·h) SV 170 155 [99] Mesoporous Co-Mn 80 ppm HCHO, 36000 h-1 SV, 50% relative humidity — [100] Mesoporous Co3O4-CeO2 200 ppm HCHO, 22 vol% O2, 55000 h-1 SV — [75]

Table 2 Overview of catalytic activities of transition-metal oxides in HCHO oxidation.

In 2001, Sekine et al. ^[81] developed a cardboard-like air purification material composed mainly of activated carbon particles and MnO_x. The material was tested by many families in Japan. This material reduced indoor HCHO concentrations from 0.21 to 0.04 ppm, and desorbed HCHO molecules from indoor building materials. In 2002, Sekine ^[82] also reported catalytic HCHO oxidations using various metal oxides such as Ag₂O, PdO, CoO, CuO, ZnO, Fe₂O₃, La₂O₃, V₂O₅, MnO₂, TiO₂, CeO₂, and Mn₃O₄. Apart from those with La₂O₃, ZnO, and CuO, Fe₂O₃, and V₂O₅, the HCHO conversion rates of these oxides were above 50%. MnO₂ showed the best catalytic activity. The catalytic activities of two types of MnO₂ were tested. The catalytic performance of the MnO₂ sample with a higher specific surface area was better, because a higher specific surface area is conducive to the adsorption and oxidation of gas-phase components. The MnO₂ catalyst contains multi-holes, therefore oxygen molecules can easily spread to the active sites on the surface, giving complete HCHO conversion at 100 °C.

Since 2002, most research has focused on MnO₂ catalysts for HCHO oxidation because its catalytic activity is better than those of other oxides such as CuO, Fe₂O₃, La₂O₃, V₂O₅, TiO₂, and CeO₂. MnO₂ has various morphologies, namely α, β, γ and δ types. Zhou et al. ^[83] used a hydrothermal method to prepare MnO₂ samples consisting of cryptomelane, birnessite, and ramsdellite, and monoclinic crystalline MnOOH nanorods. Of these, cryptomelane MnO₂ has the best catalytic properties. Chen et al. ^[84] prepared pyrolusite, cryptomelane, and todorokite manganese oxides with different channel structures. The catalytic activity of cryptomelane MnO₂ is better than those of the others (Fig. 7). The effect of pore channels in HCHO oxidation is stronger than those of crystallinity, specific surface area, reducibility, and Mn oxidation state. Cryptomelane has the best catalytic activity mainly because the HCHO molecules and pore channels are of similar size. It has been reported in Ref. ^{[83, 84]} that the crystalline structure of a catalyst determines its catalytic activity.

Fig. 7. Complete oxidation of HCHO on pyrolusite, cryptomelane, and todorokite manganese oxides [84].

Experimental and theoretical results have shown that the catalytic oxidation activity is affected by factors such as structure, specific surface area, active oxygen species, and active phase ^{[85, 86, 87]}. Most metal oxides synthesized by traditional precipitation methods have relatively small specific surface areas, fewer exposed active sites or oxygen species, and poor low-temperature reducibility. These limitations are a drawback in catalytic oxidation of HCHO. Properties such as a large specific surface area, good low-temperature reducibility, and large number of active sites enhance the catalytic ability. Materials containing the same components, but with different morphologies and structures, have different functions. Researchers are now trying to change the structures and morphologies of metal oxides to control the factors that influence their catalytic activities.

MnO₂ synthesized with various morphologies such as rods, spheres, particles, and porous have enhanced catalytic properties. Yu et al. ^[69] prepared MnO₂ materials with cocoon, sea-urchin, and nest morphologies (Fig. 8). Nest-like MnO₂ has the best HCHO oxidation activity because of its hole structure, which is beneficial for the adsorption and transformation of HCHO. The MnO₂ nests have diameters in the range 1-3 μm, and consist of nanorods of width 10-50 nm and length 0.2-1.5 μm. Chen et al. ^[88] reported that hollow K_xMnO₂ nanospheres gave good catalytic performance in low-temperature HCHO oxidation, with complete conversion at 80 °C. The catalytic activity of this material is superior to those of OMS-2, traditional MnO_x powder, and Mn-Pd/Al₂O₃. The structure and HCHO conversion data for the catalyst are shown in Fig. 9. Tang et al. ^[89] reported that the catalytic performance of OMS-2 nanorods with a molecular sieve structure consisting of octahedral Mn was better than that of traditional MnO_x. Tian et al. ^[90] prepared K-OMS-2 nanoparticles with an octahedral Mn structure; its HCHO oxidation activity was superior to that of OMS-2 nanorods. This group ^[91] also prepared MnO_x catalysts with a birnessite structure; they had excellent catalytic activities in HCHO oxidation because of the special topography, large surface area, and good surface reducibility.

Fig. 8. SEM (a, c, e) and TEM (b, d, f) images of three MnO2 nanostructures: (a, b) cocoon-like MnO2, (c, d) sea-urchin-like MnO2, (e, f) nest-likeMnO2 [69]

Fig. 9. SEM and TEM images of honeycomb KxMnO2 nanospheres (a,b) and hollow KxMnO2 nanospheres (c,d); HCHO conversions over hollow KxMnO2 nanospheres (a) and honeycomb KxMnO2 nanospheres (b) [88]

Morphological changes can improve the factors that influence catalytic activity, such as the specific surface area, low-temperature reducibility, and active phase. Research has not been confined to MnO₂ materials with various structures and morphologies. Other transition-metal oxides (Cr₂O₃ and Co₃O₄) and a rare-earth oxide (CeO₂) have been studied. Xia et al. ^[92] reported the use of 3D ordered mesoporous Cr₂O₃ in catalytic oxidation of HCHO. The TEM images in Fig. 10 show the mesoporous structure of the catalyst. Mesoporous Cr₂O₃ calcined at 400 °C has the best HCHO oxidation activity, and complete HCHO conversion is achieved at 130 °C with 500 ppm HCHO and a GHSV of 30000 mL/(g·h) (Fig. 10). The activation energy is 45.6 kJ/mol. The specific surface area of mesoporous Cr₂O₃ (69-124 m²/g) is larger than that of traditional bulk Cr₂O₃; it also has better low-temperature reducibility, which enhances its catalytic properties. The catalytic activity is also related to the catalyst preparation process and the template quality. Zhang et al. ^[72] reported a 3DOM CeO₂ catalyst with a controllable pore size. The catalytic activity in HCHO oxidation of a catalyst with a pore diameter of 80 nm is better than that of one with a pore diameter of 130 nm; it has a larger surface area and can completely convert HCHO to CO₂ and H₂O at 355 °C.

Fig. 10. TEM images (a-d) and HCHO conversions (e) over 3D Cr2O3 catalyst [92].

Our group used SBA-15 and KIT-6 as hard templates to prepare 2D Co₃O₄ and 3D Co₃O₄ catalysts ^[42]. The materials have the structural characteristics of their templates, and contain 2D hexagonal (p6mm) pore channel structures and 3D symmetrical cubic (ia3d) structures, respectively. These materials all have the crystalline cobalt oxide spinel structure. The 3D Co₃O₄ catalyst gave 100% HCHO conversion at 130 °C and a space velocity of 30000 mL/(g·h). The 2D Co₃O₄ catalyst oxidized HCHO completely at 140 °C under the same space velocity conditions (Fig. 11). The difference between the activities of these catalysts arises from the channel structure of the mesoporous Co₃O₄ prepared using the hard template method; it has a high specific surface area and a large number of surface-active species, enabling reactants to diffuse and undergo surface reactions. The Brunauer-Emmett-Teller specific surface areas of 2D Co₃O₄ and 3D Co₃O₄ are 43.3 and 85.9 m²/g, respectively; these values are much larger than that of traditional nano-Co₃O₄ (28.1 m²/g). A catalyst with a larger specific surface area exposes more active facets during oxidation. TEM images show that the active faces of 3D Co₃O₄ are (220) crystal planes (Fig. 11), which mainly consist of abundant Co³⁺ ions, which provide sufficient active sites for oxidation. Large numbers of Co³⁺ ions are exposed on its surface, further enhancing the catalytic ability. X-ray photoelectron spectroscopy shows that greater quantities of Co³⁺ ions are present on the surface of 3D Co₃O₄ than on that of 2D Co₃O₄. A larger number of Co³⁺ species can increase the number of anionic defects, and this promotes adsorption and activation of oxygen in the gas phase. Oxygen temperature-programmed desorption shows that mesoporous Co₃O₄, particularly 3D Co₃O₄, has a large number of surface-active oxygen species, which are easily stripped off and quickly filled by gas-phase oxygen. This improves the oxidation capacity. Larger surface areas, more abundant surface-active oxygen species, and a larger amount Co³⁺ active species are responsible for the improved oxidation activity. Our group also prepared CeO₂ nanospheres ^[79]. The catalytic activity of CeO₂ nanospheres in HCHO oxidation is better than that of traditional bulk CeO₂ nanoparticles (Fig. 6). The HCHO conversion rate with CeO₂ nanospheres is 85% at 200 °C, whereas that of bulk CeO₂ nanoparticles is only 40%. This is because CeO₂ nanospheres have a larger specific surface area (148.6 m²/g) than bulk CeO₂ (57.8 m²/g). A morphological change can increase the specific surface area or affect other factors that influence the catalytic activity in HCHO oxidation, and can therefore enhance the oxidation performance. Our group ^{[93, 94]} reported that the catalytic activity of a 3D ordered mesoporous MnO₂ catalyst (T₁₀₀ < 140 °C) in HCHO and ethanol oxidation is better than that of MnO₂ nanorods because it contains more Mn⁴⁺ ions as the active phase and has a higher specific surface area.

Fig. 11. TEM images (a-d) and HCHO conversions (e) over 3D Co3O4 catalyst [42].

Table 2 shows that single metal oxide catalysts such as MnO₂ nanorods, cryptomelane nanospheres, and mesorporous MnO₂, Co₃O₄, and Cr₂O₃ have good catalytic activities in HCHO oxidation, and their T₅₀ and T₁₀₀ HCHO conversions are less than or equal to 110 and 140 °C, respectively. Their catalytic activities are superior to those of traditional metal oxide catalysts synthesized using precipitation methods. This is because these catalysts have special morphologies and structures and high surface areas, which promote HCHO activation on the catalyst surface. Morphological changes can increase the surface area and, more importantly, increase the low-temperature reducibility, and numbers of active cations and surface-active oxygen species, and therefore increase the oxidation ability of the catalyst. Single metal oxide catalysts with special morphologies have a range of potential applications because they are cheap and give good catalytic performances.

The development of transition-metal oxide catalysts that can completely convert HCHO at room temperature is desirable. Some single metal oxides cannot convert HCHO at low temperatures; other elements such as Ce, Sn, Zr, and Cu have been added to manganese and cobalt oxides to prepare Mn- and Co-based catalysts.

Tang et al. ^[62] used a coprecipitation method to prepare a MnO_x-CeO₂ mixed oxide, and investigated the effect of the Mn/(Mn + Ce) molar ratio on its catalytic performance in HCHO oxidation. The best catalytic activity was obtained when the Mn/Ce mole ratio was 1:1. HCHO conversion of 90% was achieved at 90 °C with a 580 ppm feed concentration and space velocity of 30000 mL/(g·h). Formation of a solid solution and improved reducibility are responsible for the good catalytic performance. Tang et al. ^[95] also reported Mn-Ce catalysts. A MnO_x-CeO₂ sample prepared using a modified coprecipitation method completely converted HCHO at 100 °C at 580 ppm HCHO and a GHSV of 21000 mL/(g·h) (Fig. 12). The large number of Mn⁴⁺ ions and surface lattice oxygen species are responsible for the excellent catalytic activity in HCHO oxidation. Liu et al. ^[96] also prepared a MnO_x-CeO₂ catalyst. The average oxidation state of Mn species on the catalyst surface was higher than that of pure MnO_x. An improved catalytic performance in HCHO oxidation was achieved because of activation of surface lattice oxygen species. Wen et al. ^[97] reported the preparation of MnO_x-SnO₂ catalysts using a coprecipitation method. These gave good catalytic performances in HCHO oxidation because of the presence of high-oxidation-state Mn species and the formation of a Mn-Sn solid solution. Pei et al. ^[98] reported the performance and kinetics of a CuO/MnO₂ catalyst for indoor HCHO removal. The experimental data for HCHO oxidation over the CuO/MnO₂ catalyst fitted the traditional Langmuir-Hinshelwood model.

Fig. 12. HCHO conversions over various MnOx-CeO2 catalysts [95].

Bai et al. ^[99] used cyclodextrin (CD) to prepare a Co₃O₄/Zr catalyst. The catalytic performance of the material containing CD and Co in a molar ratio of 0.1:1 was better than that of a catalyst directly prepared from cobalt nitrate; this is attributed to stronger interactions between Co and the carrier and better reducibility. Complete HCHO conversion was achieved over the Co-CD_0.1/Zr catalyst at 180 °C with 150 ppm HCHO and a GHSV of 30000 mL/(g·h) (Fig. 13). Wang et al. ^[100] prepared a 3D ordered mesoporous Co-Mn mixed oxide. This catalyst removed HCHO efficiently because of its large specific surface area and specific pore structure.

Fig. 13. HCHO conversions over various Co-CDx/Zr catalysts [99].

Table 2 shows that MnO_x-CeO₂ composite metal oxide catalysts ^{[62, 95, 96]} have good catalytic activities (T₅₀ < 100 °C). This is because strong interactions between MnO_x and CeO₂ change the number of surface-active oxygen species and the active phase. Ma et al. ^[75] prepared 2D ordered mesoporous Co₃O₄-CeO₂. Under the same conditions, the HCHO conversion (20.3%) over 2D Co₃O₄ was higher than that (13.2%) over 2D Co₃O₄-CeO₂. This indicates that the addition Ce does not increase the catalytic activity, probably because a solid solution is not formed. Few composite metal oxide catalysts have been reported; therefore, syntheses of composite metal oxide catalysts with special morphologies for lower-temperature catalytic activity need to be developed.

The catalytic activities of metal and transition-metal oxide catalytic systems in HCHO oxidation are affected by factors such as the preparation method, morphology and structure, specific surface area, active sites, low-temperature reducibility, and surface active oxygen species; these effects are inter-related. In addition, HCHO conversion by catalysts is influenced by experimental parameters such as the water vapor content, initial HCHO concentration, and space velocity.

Catalysts or supports with different morphologies or higher surface areas can be obtained by using different preparation methods. The catalytic activity in HCHO oxidation varies depending on important factors such as active sites, surface-active oxygen species, and low-temperature reducibility. Changing the method used to prepare catalysts or supports is therefore an efficient way to obtain excellent catalytic activities for HCHO oxidation.

Zhou et al. ^[83] used KMnO₄ and oleic acid for the hydrothermal synthesis of MnO₂ samples with different crystalline structures (cryptomelane, birnessite, ramsdellite, and monoclinic crystals) and morphologies (nanosheets, nanospheres, and nanorods). They examined the effects of hydrothermal temperature, pH, and anion (PO₄^3-, SO₄^2- and NO₃^-) on the morphology and structure of MnO₂. The preparation method is shown in Fig. 14. The morphologies and structures of samples prepared hydrothermally at 80, 120, and 180 °C are δ-MnO₂ nanosheets, flower-like δ-MnO₂, and γ-MnOOH nanorods, respectively. The pH affects the size of cryptomelane α-MnO₂ nanorods. The samples prepared using PO₄^3-, SO₄^2-, and NO₃^- anions are flower-like δ-MnO₂, α-MnO₂ nanorods, and ramsdellite MnO₂, respectively. The results show that the crystalline structure of the catalyst significantly affects HCHO oxidation. Below 120 °C, the order of the catalytic activities in HCHO oxidation is cryptomelane MnO₂ > birnessite MnO₂ > ramsdellite MnO₂ > monoclinic MnOOH. The high activity of cryptomelane MnO₂ is attributed to the specific tunnel size of 0.46 × 0.46 nm², and large numbers of defects and surface hydroxyl groups. Tang et al. ^[95] prepared MnO_x-CeO₂ catalysts using sol-gel, coprecipitation, and modified coprecipitation methods. The MnO_x-CeO₂ sample prepared using the modified coprecipitation method gives the best catalytic performance in HCHO oxidation because it has more Mn⁴⁺ ions and surface lattice oxygen species. The formation of a solid-solution crystalline structure and improved reducibility also contribute to excellent oxidation activity. The Na-Pt/TiO₂ catalyst prepared by Zhang et al. ^[46] has the best activity in HCHO oxidation because of the addition of Na⁺ ions as auxiliaries during preparation. The addition of auxiliaries clearly increases the catalytic ability of Pt/TiO₂. HCHO at 600 ppm can be completely converted to CO₂ and H₂O at room temperature and a GHSV of 300000 h^-1 because of the large number of surface hydroxyl groups as surface-active oxygen species and better low-temperature reducibility. Our group ^[42] prepared mesoporous Co₃O₄ and bulk Co₃O₄ nanoparticles using hard template and precipitation methods, respectively. Mesoporous Co₃O₄ has better catalytic activity in HCHO oxidation, because the preparation method produces mesoporous Co₃O₄ with a special structure and morphology and an increased surface area; this enhances the active phase and low-temperature reducibility.

Fig. 14. Schematic diagram of growth of different manganese oxide nanostructures under various conditions [83].

Table 1 and 2 confirm that the preparation method affects the catalytic activities of many catalysts. Transition-metal oxides with special morphologies and Ag or Au catalysts with such oxides as supports have promising low-temperature catalytic activities in HCHO oxidation because of the preparation methods used. However, catalyst preparation involves various factors such as choice of precipitant, the deposition rate, hydrothermal time and temperature, pH, template type, calcination temperature, auxiliaries, and choice of active components and supports. Much effort has been expended on the preparation of catalytic materials, but complete elimination of HCHO at room temperature has not yet been achieved. The experimental procedures for synthesizing materials is difficult and complex. One of the main aims in catalysis research is therefore the preparation of highly efficient catalysts using different methods.

The morphologies and structures of catalysts depend on the preparation methods. Different morphologies and structure result in different surface areas, active phases or facets, reducibilities, and active oxygen species, and these affect the catalytic activities.

Our group ^{[93, 94]} prepared 3D ordered mesoporous MnO₂, and α- and β-MnO₂ nanorod catalysts using hard template and hydrothermal methods, respectively. 3D MnO₂ consists of a β-MnO₂ crystalline phase corresponding to pyrolusite with a rutile structure. It has (211) and (332) low-angle diffraction peaks, which are characteristic of 3D ordered mesoporous structures with cubic symmetry (ia3d). The unit cell parameter of 3D MnO₂, calculated from the (211) diffraction, is 20.85 nm, and the unit cell parameter of the KIT-6 template is 21.63 nm. The two numbers are nearly equal, which confirms that 3D MnO₂ successfully replicates the 3D mesoporous structure of the KIT-6 template. Fig. 15 shows the ordered cubic (ia3d) mesoporous structure. The lattice dimension (19.66 nm), which was determined using TEM, matches the value calculated from the d₂₁₁ spacing in the X-ray diffraction pattern. The electron diffraction pattern shows that 3D MnO₂ has a disordered polycrystalline wall. 3D MnO₂ also exposes the (110) and (101) crystal planes. α- and β-MnO₂ nanorods are non-porous materials (Fig. 15) with low surface areas and cryptomelane and pyrolusite, structures, respectively. 3D MnO₂ gives a better catalytic performance, mainly because it has more surface Mn⁴⁺ ions as active sites and a higher surface area, as a result of its structure and morphology. Manganese oxides with pyrolusite, cryptomelane, and todorokite structures with different channel structures were also prepared (Fig. 16). Cryptomelane MnO₂ has the best catalytic activity, mainly because its channel size matches that of HCHO molecules, rather than because of its surface area and reducibility ^[84]. The order of the catalytic activities of these MnO₂ catalysts is cryptomelane (α-MnO₂) > todorokite (δ-MnO₂) > pyrolusite (β-MnO₂), because of their different crystalline structures. If the crystalline structure is unchanged but the morphology of pyrolusite MnO₂ is changed to mesoporous, the catalytic activity order changes to pyrolusite (3D mesoporous β-MnO₂) > cryptomelane (α-MnO₂ nanorods) > pyrolusite (β-MnO₂ nanorods).

Fig. 15. TEM images of 3D MnO2 (a, b),α-MnO2 nanorods (c), and β-MnO2 nanorods (d) [94].

Fig. 16. Crystal structure models of three manganese oxides with different square tunnel sizes [84].

The structure and morphology clearly affect the catalytic activity in HCHO oxidation. There are many papers on the effects of morphology and structure on the catalytic activities of materials such as nest-like MnO₂ ^[69], hollow K_xMnO₂ ^[88], OMS-2 ^{[89, 90]}, 3DOM CeO₂ ^[72], 3D Cr₂O₃ ^[92], 3D Co₃O₄ ^[42], and CeO₂ nanospheres ^[79].

A higher surface area exposes and disperses more noble-metal active sites and activates more HCHO molecules, and this can improve the catalytic activity in HCHO oxidation. This is why TiO₂, Al₂O₃, and molecular sieves are commonly used as supports for noble-metal catalysts. Relatively cheap Au or Ag loaded on metal oxide supports with high surface areas have good potential as catalysts. Li et al. ^[65] reported that Ag/CeO₂ catalysts with different surface areas showed different catalytic activities in HCHO oxidation (Fig. 17). A higher surface area favors the formation of Au species with high oxidation states and Au_xCe_1-xO_2-δ solid solutions.

Fig. 17. Catalytic oxidation of HCHO over Au/CeO2 catalysts [65].

Higher surface areas can also improve the catalytic activities of metal oxide catalysts by enhancing factors such as the reducibility, active phase, and oxygen species. Dai’s group ^[92] prepared 3D mesoporous Cr₂O₃ with high surface areas (Table 3). 3D Cr₂O₃ has better catalytic activity than bulk Cr₂O₃, which has a smaller surface area (5 m²/g). Zhang et al. ^[72] prepared 3DOM CeO₂. The catalytic activity of 3DOM CeO₂-80 in HCHO oxidation is better than that of 3DOM CeO₂-100, because 3DOM CeO₂-80 has a higher surface area. Our group reported that 3D Co₃O₄ ^[42], 3D MnO₂ ^[94], and CeO₂ nanosphere ^[79] catalysts gave good catalytic performances because of their low-temperature reducibilities and large numbers of metal cations, resulting from their high surface areas. Some studies of oxidation of other VOCs have also shown that a higher surface area improves the catalytic activity ^{[101, 102]}.

Table 3 Textural properties of KIT-6 and Cr2O3 catalysts [92] Sample Surface area (m2/g) Average pore diameter (nm) Pore volume (cm3/g) KIT-6 790 6.1 0.95 bulk-Cr 5 — — meso-Cr-200 91 7.7 0.13 meso-Cr-300 98 7.6 0.14 meso-Cr-400 124 7.9 0.21 meso-Cr¢-400 69 8.2 0.12 meso-Cr-500 109 8.9 0.16

Table 3 Textural properties of KIT-6 and Cr2O3 catalysts [92]

However, the surface area is not the decisive factor in catalytic oxidation. Sometimes a higher surface area does not result in a better catalytic performance. For instance, Yu et al. ^[69] prepared MnO₂ catalysts with higher surface areas. Specific surface areas of cocoon-like MnO₂, sea-urchin-like MnO₂, and nest-like MnO₂ are 247.6, 62.3 and 56.9 m²/g, respectively. Fig. 18 shows that nest-like MnO₂ has the smallest surface area and the best catalytic activity; this is because the channels of width 0.6 nm are suitable for adsorption of HCHO molecules. Although cocoon-like MnO₂ has the highest surface area, its low crystallinity prevents the formation of micropores.

Fig. 18. Catalytic performances in HCHO oxidation of (a) cocoon-like MnO2, (b) sea-urchin-like MnO2, and (c) nest-like MnO2 [69].

The active sites are one of the decisive factors in determining the catalytic activity. A larger number of active sites can directly improve the oxidation abilities of catalysts, but this is a complex and abstract topic. In HCHO oxidation, noble metals (Pt, Pd, Au, and Ag) often provide better active sites. Most catalysts with Pt and Pd as active sites have excellent catalytic performances in the oxidation of HCHO and other VOCs because of the stronger ability of these metal atoms to activate O-O and C-H bonds ^{[103, 104]}. The active sites of transition-metal oxide catalyst are usually provided by metal ions with high oxidation states. The activation ability of a noble metal is much stronger than those of high-valence transition-metal cations such as Mn⁴⁺, Co³⁺, and Cr⁶⁺. This is why addition of a noble metal can result in complete HCHO conversion at lower temperatures under normal circumstances.

In the Refs. [52, 53, 70, 78, 93], active sites are often defined as the number of noble-metal atoms or metal cations with high valence states on the catalyst surface. Generally, dispersion, which is related to the noble-metal particle size and loading, determines the number of active sites on a supported noble-metal catalyst. For example, Huang et al. ^[52] reported that a PdO-DP catalyst achieved higher HCHO conversion than a PdO-IMP catalyst, because of its higher Pd dispersion, which results in more active sites. The Pd-DP-NaBH₄ catalyst has a higher TOF because of its higher Pd dispersion, Pd metallic state, and smaller particle size. Huang et al. ^[70] prepared a catalyst with stable single-atom Ag chains in HMO tunnels. Ag atoms replace K⁺ ions in the channel structure of Mn octahedral sieves. The single Ag atoms at the openings of the tunnels of Ag-HMO nanorods are defined as active sites. The Ag-HMO catalyst has a better catalytic activity than a Ag/HMO catalyst with supported Ag nanoparticles, because the higher Ag dispersion increases the number of active sites. On transition-metal oxide catalysts, surface metal cations with high valence states determine the number of active sites. Tang et al. ^[95] reported a MnO_x-CeO₂ sample that had good catalytic activity because of the large number of surface Mn⁴⁺ ions and lattice oxygen species. Our group reported that more metal cations clearly improved the HCHO oxidation activities on 3D Co₃O₄ ^[42] and 3D MnO₂ ^[94] catalysts.

Low-temperature reducibility is an important factor in oxidation reactions. Better reducibility means that oxygen species are easily activated or can easily migrate to the catalyst surface. It has been reported that better catalytic activity is related to lower-temperature reducibility. He’s group reported that the reducibility of a Na-Pt/TiO₂ catalyst correlated with its activity ^[46]. Fig. 19 shows that Na addition shifts the reduction peak to lower temperature. Tang’s group prepared Ag/MnO_x-CeO₂ ^[63], which has good catalytic activity because of its low-temperature reducibility. Ag/MnO_x-CeO₂ shows a reduction peak at 293 °C. The presence of Ag shifts the reduction peak to lower temperature compared with that for MnO_x-CeO₂, because of interactions between Ag and MnO_x-CeO₂. Dai’s group reported that 3D Cr₂O₃ calcined at 400 °C had the best catalytic activity in HCHO oxidation among a group of Cr catalysts calcined at different temperatures ^[92]. The initial H₂ consumption rate was used to evaluate the catalyst reducibilities. The order of the initial H₂ consumption rates is Cr-400 > Cr-300 > Cr-500 > Cr-200 > bulk Cr, which indicates that 3D Cr-400 has the best reducibility, which correlates with the catalytic activities. Our group also reported that a K-Ag/Co₃O₄ catalyst has the best activity and low-temperature reducibility of all samples examined ^[78]. The addition of K⁺ ions shifts the reduction peak to lower temperature.

Fig. 19. H2-temperature-programmed reduction profiles of x wt% Na-1% Pt/TiO2 (x = 0, 1, and 2) [46]

A large number of surface oxygen species improves the catalytic activity. The adsorption, activation, and migration of oxygen during the oxidation reaction is complex. The consumption and supply of oxygen species occur continuously on the catalyst surface. Generally, a catalyst with abundant surface oxygen species has better activity than one with few such species in the catalytic oxidation of HCHO.

Our group loaded Ag nanoparticles on 3D Co₃O₄ ^[78]. Ag/Co₃O₄ has more surface lattice oxygen species because of the presence of Ag, and has better catalytic activity than 3D Co₃O₄. The addition of K⁺ ions to Ag/Co₃O₄ gave K-Ag/Co₃O₄ has a large number of surface hydroxyls as surface adsorbed oxygen species, because of the presence of K⁺ ions, and shows better activity than Ag/Co₃O₄. Surface adsorption and lattice oxygen species on K-Ag/Co₃O₄ are involved in the activation and migration of oxygen, and this increases the oxidation activity in HCHO. Zhang et al. ^[46] added Na⁺ ions to Pt/TiO₂, which provided abundant surface hydroxyls on the Na-Pt/TiO₂ catalyst. Na-Pt/TiO₂ shows the best activity in HCHO oxidation. Surface hydroxyls as active oxygen species directly change the reaction route of HCHO oxidation. Different preparation methods lead to different surface oxygen species. Large amounts of surface chemisorbed oxygen form on Ag/CeO₂ ^[79] and 3D Co₃O₄ ^[42], and this improves their catalytic activity in HCHO oxidation.

Different experimental parameters such as water vapor content, initial HCHO concentration, and space velocity, can lead to different HCHO conversions. Generally, catalytic conversion of HCHO decreases with increasing initial HCHO concentration and GHSV. Huang et al. ^[52] examined the effect of initial HCHO concentration in the range 5-30 ppm. The steady conversions of HCHO were 100% at 5 ppm, 99.1% at 10 ppm, 98.2% at 20 ppm, and 95.8% at 30 ppm. As anticipated, the HCHO removal efficiency decreased with increasing initial HCHO concentration. The effect of GHSV on HCHO oxidation was investigated using 0.1% Pt-TiO₂ in the range 40000-240000 h^-1. The steady conversions of HCHO were 100% at 40000 h^-1, 99.1% at 80000 h^-1, 97.8% at 160000 h^-1, and 90.5% at 240000 h^-1. The HCHO conversion decreased with increasing GHSV. Tang et al. ^[62] investigated the effect of HCHO concentration. They achieved 100% conversion of HCHO at HCHO concentrations less than 100 ppm; the conversion decreased slightly with increasing HCHO concentration. The HCHO conversion was 50% at an HCHO concentration of 580 ppm. The HCHO conversion decreased with increasing feed concentration. Xia et al. ^[92] examined the effects of GHSV on HCHO, acetone, and methanol oxidation in the range 10000-60000 h^-1. The results confirm that conversion over 3D Cr₂O₃ decreased with increasing of GHSV.

For the catalytic oxidation of other VOCs such as ethanol ^{[93, 105]}, CH₄ ^[101], and toluene ^[106], the addition of water vapor shifts complete conversion to higher temperatures. The VOC conversion decreases with increasing water vapor content. However, the presence of water vapor improves HCHO conversion. Huang et al. ^[52] studied the effect of moisture. HCHO oxidation was promoted by water vapor up to 50% humidity. Nearly 100% HCHO conversion was achieved at humidities of 25% and 50%, but the conversion was only 45.2% at the stable stage at 0% humidity. The HCHO conversion decreased slightly at humidities higher than 50% because of coverage of active centers by water, but the removal efficiency was still as high as 95.6% at 97.5% humidity. The presence of moisture did not inhibit, but enhanced, HCHO oxidation at room temperature.

The catalytic activities in HCHO oxidation of different catalysts were compared under the same conditions. Changes in the GHSV and initial HCHO concentration hinder comparison of the oxidation activities of various catalysts. Normalized rates, i.e., mol/(m²·h) or mol/(m²·s), and TOFs (s^-1 or h^-1) should therefore be used calculate HCHO oxidation activities. In real situations, water vapor is often present in indoor environments and vehicle exhausts. It is therefore necessary to examine the effect of moisture on catalytic activity.

Research on the mechanism of HCHO oxidation is a sophisticated process. Several mechanisms have been proposed based on the Mars-van Krevelen mechanism. He’s group ^[45] reported the reaction mechanism on TiO₂-supported noble-metals catalysts (Fig. 20). Dioxymethylene, formate, and adsorbed CO are important reaction intermediates in HCHO oxidation. The decomposition of surface formate species to adsorbed CO on the catalyst is the rate-determining step for catalytic oxidation of HCHO. HCHO is oxidized to surface dioxymethylene species, and then formate species. The surface formate species decompose to adsorbed CO species and H₂O, and then the CO species react with O₂ to produce gas-phase CO₂. Pt/TiO₂ shows high activity in HCHO oxidation because decomposition of formate species is easier on this catalyst than on supported catalysts containing Rh, Pd, and Au. The same group reported that the reaction mechanism over a Na-Pt/TiO₂ catalyst differed from that on a Pt/TiO₂ catalyst ^[46]. Na-Pt/TiO₂ has the best catalytic activity in HCHO oxidation because of the surface hydroxyl groups formed by Na⁺ ions; the reaction mechanism differs from that over Pt/TiO₂ because of the addition of Na. The HCHO oxidation reaction on Na-free Pt/TiO₂ follows the formate decomposition route (HCHO → CHOO^- → CO → CO₂), with formate decomposition to CO being the rate-determining step. However, HCHO oxidation over 2% Na-1% Pt/TiO₂ follows a different pathway, HCHO → CHOO^- + •OH → CO₂ + H₂O. The reaction between surface hydroxyls and formate is easier than decomposition of formate to CO followed by CO oxidation. The process in which formate species on the catalyst surface directly react with surface hydroxyl groups to form CO₂ and H₂O becomes the rate-determining step. Nie et al. ^[47] reported the same reaction mechanism for HCHO oxidation on the Na-Pt/TiO₂ catalyst.

Fig. 20. Reaction scheme for catalytic oxidation of HCHO on TiO2-supported Pt, Rh, Pd, and Au catalysts [45]

Our group ^[78] reported the reaction mechanism for HCHO oxidation on a 3D K-Ag/Co₃O₄ catalyst (Fig. 21). K-Ag/Co₃O₄ has a large number of Ag (111) crystal facets, surface lattice oxygen (O^2-) species, and Co³⁺ cations because of the addition of K⁺ ions; these are all involved in the oxidation reaction. K⁺ ion addition changes the reaction route of HCHO oxidation on 3D K-Ag/Co₃O₄ catalyst, as in the case of 2% Na-1%Pt/TiO₂ catalyst ^[46]. Research shows that more Co³⁺ cations can increase the oxygen vacancy density; these vacancies result from anionic structural defects. The oxygen vacancies directly participate in the adsorption, activation, and migration of oxygen. The Ag (111) planes are active faces in HCHO oxidation and can activate the O^2- species of 3D Co₃O₄. The O^2- species can also improve the reactivity of the Ag (111) surfaces, enhance the breakage of H₂, O₂, and NO bonds, and strengthen the bonding of H, O, N, and C atoms to the Ag (111) surface. In the reaction, the active O^2- species around Ag are directly depleted and replenished by the Co₃O₄ support, which acts as an oxygen reservoir. The activation and migration of oxygen species at the oxygen vacancies depend on the Co³⁺/Co²⁺ and Ag⁺/Ag⁰ redox cycles after the O^2- species are consumed (Fig. 21, yellow box). The redox cycle is possibly the same as the Mn⁴⁺/Mn³⁺ and Ce⁴⁺/Ce³⁺ redox cycles reported by Tang et al. ^[62]. The increase in the number of Co³⁺ ions favors the formation of oxygen vacancies, which can enhance the Co³⁺/Co²⁺ and Ag⁺/Ag⁰ redox cycles. In addition, the addition of K⁺ ions results in the presence on the K-Ag/Co₃O₄ catalyst surface of adsorbed oxygen species, in the form of hydroxyl species. Surface hydroxyl species play a critical role in the reaction path of HCHO oxidation. A surface hydroxyl can immediately react with formate species on the K-Ag/Co₃O₄ surface to form a molecule of CO₂ and H₂O. This is similar to the reaction path for the Na-Pt/TiO₂ catalyst ^[46]. The difference is that O^2- species at the perimeter of Ag in K-Ag/Co₃O₄ participate in HCHO oxidation because of strong interactions between Ag with Co and anionic lattice defects. The TOFs of K-Ag/Co₃O₄, especially 1.7% K-Ag/Co₃O₄, are much higher than that of Ag/Co₃O₄. It is concluded that at low temperature (< 80 °C), the catalytic activity of K-Ag/Co₃O₄ in HCHO oxidation largely depends on surface •OH species at the perimeters of Ag (111) facets; at higher temperatures (>80 °C), the surface •OH species are consumed and replaced quickly, and their supply relies on the migration of O^2- species from the 3D Co₃O₄ support. The reaction pathway of HCHO oxidation on K-Ag/Co₃O₄ is therefore HCHO → CHOO^- + •OH → CO₂ + H₂O. For Ag/Co₃O₄, the reaction pathway is HCHO → CHOO^- → CO + O* → CO₂. The O^2- species play an important role on the oxidation reaction and active O* species are often involved in complex migration though oxygen vacancies. The TOFs of Ag/Co₃O₄ show that the catalytic activity depends on the O^2- species around Ag nanoparticles at low temperature (<90 °C), and on the O^2- species at the perimeter of Ag, which migrate from the 3D Co₃O₄ support at higher temperatures (>90 °C).

Fig. 21. Reaction pathway on K-Ag/Co3O4 catalyst [78].

Liu et al. ^[73] reported the mechanism of HCHO oxidation on a 3DOM Au/CeO₂ catalyst (Fig. 22). HCHO oxidation occurs via two processes, catalyzed by ionic Au³⁺ and metallic Au⁰, respectively; Au³⁺ shows higher catalytic activity. The CeO₂ support in contact with Au³⁺ may be partly reduced to Ce₂O₃ when Au³⁺ ions are formed on the surface of the 3DOM Au/CeO₂ catalyst. After adsorption of HCHO molecules on the surface of the CeO₂ support, active oxygen may be transferred from Au₂O₃ to HCHO to form HCOOH and Au⁰. HCOOH is then converted to formate through interactions with the CeO₂ support, accompanied by loss of •H to form H₂O with free •OH adsorbed on the surface of the CeO₂ support. Incomplete oxidation of HCOOH also occurs, which enables conversion of HCOOH to carbonate and hydrocarbonate. H₂O is generated when the lost •H interacts with free •OH. HCHO is converted to CO₂ and H₂O via this cycle. The Au³⁺ in the 3DOM Au/CeO₂ catalyst determines the efficiency of the HCHO conversion. HCHO oxidation catalyzed by metallic Au⁰ via another process is also involved, but is not the dominant process. After HCHO adsorption on the CeO₂ support, active oxygen is transferred to HCHO to form HCOOH, and then HCOOH is converted to CO₂ and H₂O, completing the oxidation process. Incomplete oxidation occurs simultaneously to form carbonate and hydrocarbonate, which may deactivate the 3DOM Au/CeO₂ catalyst. In the HCHO oxidation process, there is a balance between HCOOH adsorption and desorption, to form formate on the CeO₂ support. HCOOH can therefore be converted to CO₂ and H₂O through complete oxidation, and formate can be converted to carbonate and hydrocarbonate via incomplete oxidation; these are deposited on the surface of the CeO₂ support and deactivate the catalyst. Blockage of the active sites on 3DOM Au/CeO₂ by carbonate and hydrocarbonate is difficult, because of the macroporous structure of the catalyst.

Fig. 22. Mechanism of HCHO oxidation on 3DOM Au/CeO2 catalyst [73].

According to the literature, the addition of Na⁺ or K⁺ ions, which results in formation of hydroxyl oxygen species on the catalyst surface, can change the reaction route and increase the catalytic activity. This can be used to help design methods for catalyst preparation. Perhaps the HCHO oxidation activities of other supported noble-metal catalysts could be increased by alkali metal addition. In HCHO oxidation, abundant surface oxygen species participate in the oxidation reaction, and this enhances the catalytic activity. Surface-active oxygen species depend on the redox cycles of high- and low-valent states of metal cations, and this directly affects the adsorption, activation, and migration of oxygen species. Park et al. ^[54] reported indirect and direct reaction pathways over a 0.25 Pd/20 Mn/Bata catalyst; these are two types of redox cycle, and the cycle depends on whether Mn cations participate in the reaction at a given temperature (Fig. 23). Tang et al. ^{[62, 63]} proposed two redox cycles for HCHO oxidation over Pt/MnO_x-CeO₂ and Ag/MnO_x-CeO₂ catalysts to explain the formation of active oxygen; these are displayed in Fig. 24. HCHO oxidation over different catalysts may involve different reaction mechanisms, probably because different active oxygen species and sites can form different reaction intermediates. Formate species are important reaction intermediates in HCHO oxidation; they are present on all reported catalysts, and can produce different intermediates on different catalysts. There are three processes. (1) Formate species can decompose to adsorbed CO followed by CO oxidation on the surfaces of catalysts such as Pt/TiO₂ ^[45] and Ag/Co₃O₄ ^[78]. (2) Formate species can react with surface hydroxyls to form CO₂ and H₂O directly on the surfaces of catalysts such as Na-Pt/TiO₂ ^[46] and K-Ag/Co₃O₄ ^[78]. (3) Formate species can form carbonate and then decompose to CO₂ on the surfaces of catalysts such as 3DOM Au/CeO₂ ^[73] and 2D Au/Co₃O₄-CeO₂. The mechanism of HCHO oxidation on a 2D Au/Co₃O₄-CeO₂ catalyst was reported by Ma et al. ^[75]. As shown in Fig. 25, HCHO is oxidized to formate species, which form carbonate. The carbonate decomposes to CO₂ and OH^-. Further studies are needed to enable a deeper understanding of the mechanism of HCHO oxidation. The reaction mechanisms over more catalysts need to be clarified to obtain a more comprehensive understanding.

Fig. 23. HCHO oxidation pathway over 0.25 Pd/20 Mn/Bata catalyst [54].

Fig. 24. Redox cycles in HCHO oxidation over MnOx-CeO2 and Ag/MnOx-CeO2 [62, 63].

Fig. 25. Reaction route for HCHO oxidation over mesoporous Au/Co3O4-CeO2 [75].

HCHO is a carcinogenic and teratogenic substance, and has strong photochemical activity. It is emitted during the production of textiles, agrochemicals, sheet materials, and fine chemicals, from motor vehicle exhausts, and from various indoor decorating materials. HCHO removal is necessary to protect human health and the atmospheric environment. The removal of HCHO through catalytic oxidation is a more promising technique than adsorption methods. The key to this technique is the development of suitable catalysts. Catalytic materials for HCHO oxidation can be divided into noble-metal and transition-metal oxide systems. Noble-metal catalysts are generally prepared by loading noble metals (Pt, Pd, Au, and Ag) on various supports. Supports can be classified as common supports, traditional metal oxide supports, and metal oxide support with special morphologies. The common supports used in HCHO oxidation are usually materials such as SiO₂, Al₂O₃, TiO₂, and molecular sieves. These have large surface areas, and this is conducive to the exposure of active sites and the adsorption and diffusion of reactants and products, and can enhance the synergistic effect between the support and the active component. The catalytic HCHO oxidation activities of different noble-metal catalysts loaded on common supports decrease in the following order: Pt > Pd > Rh > Au > Ag. This type of catalyst is suitable for applications in indoor environments and factories, and for vehicle exhausts, except in terms of cost and thermostability. Na-Pt/TiO₂ ^[46], which is used in air purifiers, is the best catalyst for HCHO removal, followed by Pt/TiO₂ and Pd/TiO₂ ^[53]. The use of different supports with the same active components directly affects the activities of catalysts for HCHO oxidation. Precious metals have been loaded on metal oxide supports with the aim of obtaining the advantages of both noble-metal and metal oxide catalysts; such catalysts have better low- temperature oxidation activities and thermostabilities because of the interactions between the metal and the support. Metal oxide supports such as CeO₂, Fe₂O₃, Co₃O₄, and MnO₂, or their composites, prepared by conventional precipitation, coprecipitation, or sol-gel methods, used in HCHO oxidation are defined as traditional metal oxide supports. This type of supported Pt catalyst has excellent catalytic activity (T₅₀) at room temperature; examples are Pt/MnO_x-CeO₂ ^[62] and Pt/Fe₂O₃ ^[67]. Although supported Pt catalysts are promising for various applications, they are too expensive and their industrial production and use are limited. If the Pt loading is decreased to reduce the production cost, the reaction performance of the catalyst may be compromised. Some catalysts with traditional metal oxide supports, such as Au/CeO₂ ^[65], Ag/MnO_x-CeO₂ and Ag/CeO₂ ^[63], have better development potential, because of the strong interactions between the noble metal and the support. The choice of support for HCHO oxidation catalysts is important. The trend is not to choose Pt or Pd as the active components of catalysts of HCHO removal, and supported Au or Ag catalysts may provide suitable alternatives. Traditional metal oxide supports can be modified by using different preparation methods. Metal oxides with special morphologies and structures, such as nanorods, nanospheres, mesopores, and macropores, are mainly prepared using hydrothermal and hard template methods. Such materials may provide new supports with catalytic activities better than those of conventional bulk metal oxides prepared using precipitation methods. Few Pt catalysts have been produced using metal oxide supports with special morphologies, because 2% Na-1% Pt/TiO₂ has become the benchmark for supported Pt catalysts in air purification ^[46]. More attention has been paid to catalysts consisting of Ag and Au loaded on metal oxide supports with special morphologies. These catalysts, e.g., mesoporous 3DOM Au/CeO₂-Co₃O₄ ^[74], 2D Au/Co₃O₄-CeO₂, 2D Au/Co₃O₄ ^[75], and 3D K-Ag/Co₃O₄ ^[78], have excellent catalytic activities in room-temperature HCHO oxidation (T₅₀ = room temperature) and have good potential applications.

Transition-metal oxide catalysts have cheap and plentiful sources, and have been widely studied. Single transition-metal oxide catalysts such as MnO₂ nanorods, cryptomelane nanospheres, and mesorporous MnO₂, Co₃O₄, and Cr₂O₃ have good catalytic activities in HCHO oxidation, and their T₅₀ and T₁₀₀ HCHO conversions are less than or equal to 110 and 140 °C, respectively. Their catalytic activities are superior to those of traditional metal oxide catalysts synthesized using precipitation methods, because of their special morphologies and structures, higher surface areas, and other factors that improve the catalytic activity. Single transition-metal oxide catalysts with special morphologies have good application prospects because they are cheap and give better catalytic performances. Some metal elements (such as Ce, Sn, Cu, and Zr) can be doped into MnO_x and Co₃O₄ to prepare composite metal oxides when a single metal oxide does not give a good catalytic performance. The composite metal oxide catalyst MnO_x-CeO₂ [95,96] has excellent catalytic activity (T₅₀ < 100 °C) because the strong interactions between MnO_x and CeO₂ change the number of surface-active oxygen species and the active phase. Few composite metal oxide catalysts have been reported; therefore, the development of composite metal oxide catalysts with special morphologies for lower-temperature catalytic activities is an area for future exploration. The use of different preparation methods may provide single transition-metal oxides or composite oxide catalysts that can be used as substitutes for noble-metal catalysts, including Au and Ag, for VOC oxidation.

The catalytic activity in HCHO oxidation is affected by factors such as the preparation method, morphology and structure, specific surface area, active sites, low-temperature reducibility, and surface-active oxygen species; these factors are related to each other. Different preparation methods can provide various catalysts or supports with different morphologies or higher surface areas, which may further change some important factors such as the active sites, surface-active oxygen species, and low-temperature reducibility of the catalysts. Changing the methods used to prepare catalysts or supports is therefore an efficient way to obtain excellent catalytic activities in HCHO oxidation. Transition-metal oxides with special morphologies and their supported Ag or Au catalysts have promising low-temperature catalytic activities in HCHO oxidation because of the special preparation methods used. Catalytic HCHO conversion is also influenced by experimental parameters such as water vapor content, initial HCHO concentration, and GHSV. The activities of various HCHO oxidation catalysts were compared under the same condition. Changes in the GHSV and initial HCHO concentration are not useful in comparison of the oxidation activities of different catalysts. Normalized rates (mol/(m·h) or mol/(m·s)) and TOFs (s^-1 or h^-1) should be used to express the activity in HCHO oxidation. In real situations, water vapor is often present in indoor environments and vehicle exhausts. It is therefore necessary to examine the effects of moisture on catalytic activity.

The development of effective and low-cost catalysts for HCHO oxidation at low temperatures, even room temperature, is still an important challenge. Catalyst preparation involves various factors such as choice of precipitant, deposition rate, hydrothermal time and temperature, pH, templates, calcination temperature, auxiliaries, and choice of active components and supports. Much time has been spent on the preparation of catalytic materials, but catalysts with improved oxidation abilities are still needed. The experimental processes involved in synthesizing materials are difficult and complex. Compared with noble-metal catalysts, transition-metal oxides have better high-temperature catalytic activities, and it is very difficult to completely convert HCHO at room temperature. To control some of the factors influencing catalytic activity, such as structure, morphology, and specific surface area, metal oxides with rod-like, spherical, and porous structures, with high surface areas, have been prepared using various synthetic methods to improve the catalytic performances in HCHO oxidation. For example, the low-temperature catalytic activities of 3D Cr₂O₃ and 3D Co₃O₄, and MnO₂ nanorods and nanospheres are better than those of traditional bulk Cr₂O₃, Co₃O₄, and MnO_x. However, there is still a need for catalysts with special morphologies that can completely convert HCHO at low temperatures, even room temperature. Catalysts with carriers with different morphologies are promising for achieving complete HCHO conversion at low temperatures. A potential catalyst can be prepared by loading Pt or Pd on the support surface. However, a balance needs to be struck between the cost of precious metals and catalytic performance. Ag and Au are relatively cheap precious metals. The synthesis of catalysts consisting of Ag or Au loaded on metal oxide supports with special morphologies may be a future trend in developing catalysts supported on transition-metal oxides for complete conversion of HCHO at room temperature.

Research on the mechanism of HCHO oxidation is a sophisticated process; few mechanisms have been reported in the literature. To understand HCHO oxidation in depth, it is necessary to study the reaction mechanism further. The reaction mechanisms over more catalysts need to be reported in the future to obtain a more comprehensive understanding of HCHO oxidation. The structure-activity relationships of the catalysts also need to be further researched using various physical and chemical characterization techniques, especially in situ Raman, infrared, and near-edge X-ray absorption fine structure spectroscopies. The preparation and development of new nanocatalysts with various morphologies and structures will become a research trend in the future. Such materials can be used in the catalytic oxidation of HCHO and the catalytic oxidation of benzene series or other VOCs, and could provide techniques for decreasing VOC discharges from vehicle exhausts and industrial processes. The removal of VOCs helps to decrease of PM2.5 levels and to improve the atmospheric air quality.

This study was supported by the State Environmental Protection Key Laboratory of Sources and Control of Air Pollution Complex, and also supported by the State Key Laboratory of Environmental Criteria and Risk Assessment, Chinese Research Academy of Environmental Science.