Indoor air quality has been of increasing concern in recent years, as it directly affects health because of prolonged indoor inhabitation ^{[1, 2, 3]}. People generally spend more than 80% of time inside ^[4], where common pollutants including CO, nitrogen oxides (NO_x), and volatile organic compounds (VOCs) can adversely affect health ^[5]. Modern buildings are designed to be more airtight, to decrease energy consumption. Decreasing the intake of fresh air causes an undesirable build-up of these indoor air pollutants. Domestic appliances such as stoves and heaters are common sources of pollutants, particularly those burning coal, oil and natural gas in poorly ventilated areas ^[6]. Conventional remediation techniques include adsorption and filtration, which are expensive and inappropriate for low-concentration pollutants ^[7]. Filters without an adequate replacement can also be a source of VOCs in ventilation systems ^[8]. A different approach is required to reduce pollutant levels and maintain clean environments for good health.

Photocatalysis is a promising solution for removing indoor air pollutants. TiO₂ ^{[9, 10, 11, 12, 13, 14, 15, 16]}, bismuth titanate and strontium titanate ^{[17, 18, 19]} are attractive photocatalysts because of their strong oxidizing power, non-toxicity, and long-term photostability. TiO₂ photocatalysts are widely reported, usually in the form of highly dispersed or suspended fine particles/powders in liquid media. However, powdered TiO₂ photocatalysts are unsuitable for air purification, because particles may become respirable and cause health problems. Various studies have immobilized TiO₂ particles as thin films on rigid supports, such as glass, stainless steel, and aluminum plates ^{[20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32]}.

Coating a substrate affects its surface adsorption of reactants for photocatalysis. Photocatalytic thin films are usually coated on flat surfaces, such as honeycomb air filters. In the current study, we used three-dimensional (3D) porous ceramic foam as the coating substrate, because of its excellent hydrodynamic properties for gas passing. The ceramic foam has a versatile pore density, surface area, and chemistry. The 3D porous ceramic foam air filters have a high bed porosity, which results in the low-pressure drop required by such filters. Unlike honeycomb air filters, the 3D porous ceramic foam filters have convoluted pore structures that promote turbulence and mixing. Their open porous and reticulate structure provides favorable gas-dynamic properties, and sufficient contact between the reactant gas and catalyst surface. Porous materials are advantageous in liquid or gas phase catalysis ^{[33, 34, 35]}. Thus, highly porous materials such as ceramic foam ^{[36, 37, 38]}, porous alumina ^{[39, 40]}, porous silica ^[41], zeolite ^[42], and activated carbon ^[43] are common catalyst supports.

Depending on the method, depositing TiO₂ films on solid substrates may reduce the effective surface area of the TiO₂ photocatalyst, and thus decrease photocatalytic activity. TiO₂ thin films with a mesoporous structure exhibit high photocatalytic activity, because the structure has a high specific surface area with abundant active sites ^{[44, 45, 46, 47]}. Mesoporous TiO₂ coatings on flat glass, stainless steel, and aluminum substrates have exhibited enhanced photocatalytic efficiency in environmental purification applications ^{[16, 44, 48, 49]}. In the current study, we coated high surface area 3D porous foam filters with uniform mesoporous anatase TiO₂ thin films. The mesoporous TiO₂ coatings had a large surface area and exhibited high photocatalytic activity, and overcame the disadvantages associated with powdered TiO₂ photocatalysts. This is the first reported photocatalytic degradation of NO by mesoporous TiO₂ thin films coated on 3D porous ceramic foam air filters prepared via the reverse micellar method. NO was used as the target air pollutant because it is a major indoor air pollutant with adverse health effects. Levels of NO and NO₂ in indoor environments are normally several hundred ppb and less than a hundred ppb, respectively ^{[50, 51, 52]}.

All chemicals were of reagent grade and were purchased from Aldrich, USA.

Triton X-100 and water were added to cyclohexane to obtain 0.2 mol/L Triton X-100, at a 1.5 molar ratio of water to surfactant. After stirring for 2 h to form a microemulsion, the reverse micellar solution was mixed with titanium isopropoxide (0.4 mol/L). The alkoxide solution was stirred at room temperature for 1 h to hydrolyze and condense the titanium alkoxide. A TiO₂ sol was formed within the water droplets. The porous ceramic foam filters were then dip-coated in the TiO₂ reverse micellar solution at ambient atmosphere. The withdrawal speed was controlled at 4 mm/s. The TiO₂ gel films on the porous ceramic foam filters were calcined in air at a heating rate of 3 °C/min to 500 °C, and were then placed in a furnace at the highest temperature for different times. After calcination, transparent TiO₂ films on the porous ceramic foam filters were obtained (Fig. 1). Porous ceramic foam filters containing pore densities of 10, 20, and 30 pores per inch (ppi) were used for comparison.

Fig. 1. Images of the photocatalytic porous ceramic air filter. Inset: Higher magnification of the ceramic foam air filter showing the porous structure.

X-ray diffraction (XRD) patterns were obtained using a Bruker D8 Advance X-ray diffractometer with Cu-K_α radiation (λ = 1.54178 Å) at a scan rate of 0.05°/s, and were used to identify the phase compositions and crystallite sizes of samples. An accelerating voltage of 40 kV and an applied current of 40 mA were used. X-ray photoelectron spectroscopy (XPS) was performed using a PHI Quantum 2000 XPS System, with a monochromatic Al-K_α source and charge neutralizer. All binding energies were referenced to the C1s peak of adventitious carbon at 284.8 eV. Ultraviolet-visible (UV-Vis) diffuse reflectance spectra were obtained using a UV-Vis spectrophotometer (Cary 100 Scan Spectrophotometers, Varian, USA). The morphology and chemical composition of the thin films were examined using a scanning electron microscope (SEM, LEO 1450VP) attached to an energy-dispersive X-ray spectrometer (EDXS, Oxford Instrument). Samples were coated with Au before SEM imaging, using a sputtering thin film coating system. Brunauer-Emmett-Teller (BET) surface areas (S_BET) and pore parameters of the powder samples were determined by N₂ adsorption-desorption isotherm measurements at -196 °C, using a nitrogen adsorption apparatus (Micromeritics ASAP 2010). All samples were degassed at 180 °C prior to measurement. Pore size distributions were calculated from desorption branches of the isotherms by the Barrett-Joyner-Halenda (BJH) method using the Halsey equation. The surface roughness of the TiO₂ thin films was evaluated by atomic force microscopy (AFM, NanoScope 3a, Digital Instruments Inc., Santa Barbara, CA, USA).

The photocatalytic activities of the samples for the oxidation of NO in air were performed at ambient temperature in a continuous flow reactor (Fig. 2). The rectangular reactor was made of stainless steel and covered with Saint-Glass, and had a volume of 27.3 L (13 × 70 × 30 cm). Illumination was provided by three 14-W commercial UVA light tubes, which emit a primary wavelength at 365 nm, and the intensity was determined using a UV meter (Spectroline DRC-100X). The UV light tubes were horizontally placed outside the reactor above the sample. The TiO₂-coated porous ceramic foam filters (30 × 30 × 1 cm) were fixed within a polytetrafluoroethylene film-coated stand, and placed horizontally 5 cm below the UV lamp. The total UV intensity at 310‒400 nm incident on the filters was 540 ± 10 mW/cm², and the UV peak wavelength was 365 nm, as determined using a UV radiometer (UVX, UVP, Inc., CA, USA). The initial concentration of water vapor was 1.20 ± 0.02 vol%, and the initial temperature was 25 ± 1 °C. Stainless steel sampling ports and polytetrafluoroethylene tubing were used to connect the reactor and analytical instruments.

Fig. 2. Schematic design of the experimental setup for NOx removal test by using one photocatalytic ceramic foam air filter (30 × 30 × 1 cm).

NO gas was used as the target pollutant for the photocatalytic degradation at ambient temperature. A compressed gas cylinder provided 48 ppm NO (N₂ balance, BOC gas) with a traceable standard, from the National Institute of Stands and Technology (NIST). The NO initial concentration was diluted to 1000 or 400 ppb with an air stream supplied using a zero-air generator (Thermo Environmental Inc. Model 111). The required humidity level of the NO flow was controlled at 70% (2100 ppm) by passing the zero-air stream through a humidification chamber. The gas streams were completely pre-mixed using a gas blender, and the flow rate was controlled at 4 L/min with a mass flow controller. After the inlet and outlet concentrations of the target pollutants achieved adsorption-desorption equilibrium, the UV lamp was switched on and the reaction was initiated. The NO concentration was continuously measured using a chemiluminescence NO analyzer (Thermo Environmental Instruments Inc. Model 42c), which monitored NO, NO₂, and NO_x (where NO_x represents NO + NO₂) with a sampling rate of 0.7 L/min. The NO analyzer was calibrated after each set of experiments. Calibration was performed using a polytetrafluoroethylene sample-line filter, which was also used during analysis. The flow rate for calibration was higher than the total flow rate required by the analyzer and any other flow demand connected to the manifold. The NO analyzer was allowed to sample zero air until a stable reading was obtained. Span adjust was achieved by selecting the desired NO concentrations. The response of the NO analyzer was plotted against the corresponding NO concentrations. The experimental points were connected using a straight line and determined via linear regression techniques. A minimum R₂ value of >0.98 was achieved for all experiments.

The removal rate (%) of NO was defined as:

NO removal rate (%) = (C₀ - C) / C₀ × 100% (1)

where C₀ and C represent the NO concentrations of the feed and outlet streams, respectively. The reaction of NO with air could be ignored when performing a control experiment with or without light in the absence of photocatalyst. The photocatalytic degradation of NO is a pseudo-first-order reaction, and its kinetics may be expressed as:

ln C/C₀ = -kt (2)

where k is the apparent rate constant of pseudo-first order.

The morphologies of the TiO₂ thin films before and after coating on the porous ceramic foam filters were examined via SEM. Fig. 3(a) and 3(b) show SEM images of the selected area of the bare porous ceramic foam from different angles. The porous structure of the ceramic foam can be seen in the higher-magnification images (Fig. 3(c) and 3(d)). The original surface of the bare ceramic foam was relatively smooth (Fig. 3(c) and 3(d)) compared with the ceramic foam coated with TiO₂ (Fig. 4(c) and 4(d)). The 3D-interconnected porous networks had a large surface-to-volume ratio, so were potentially good adsorption sites for photocatalytic reaction. Chemical composition analyses using EDXS (Fig. 3(e) and 3(f)) showed that no signal corresponding to titanium (4.5 keV) was detected on the surface of the uncoated porous ceramic foam.

Fig. 3. SEM images (a-d) and EDX microanalysis spectra (e, f) of the uncoated air filters with different magnifications. No signals corresponding to Ti were detected in the EDX microanalysis spectra.

Fig. 4. SEM images (a-d) and EDX microanalysis spectra (e,f) of the TiO2-coated air filters with different magnifications. Ti signals corresponding TiO2 were detected in the EDX microanalysis spectra. The X-ray excitation energy for Ti is 4.5 keV.

Figure 4 shows SEM images of the selected areas of the TiO₂-coated porous ceramic foam. The cross-sectional view and morphology of the TiO₂ thin films coated on the ceramic foam are shown. The rigid surface of the ceramic substrate showed no obvious change after coating with TiO₂. Higher-magnification images (Fig. 4(c) and 4(d)) show that the TiO₂ was uniformly deposited on the surface of the ceramic foam substrate. The original smooth surface of the ceramic foam (Fig. 3(d)) became rough and squama-like after coating (Fig. 4(d)). To determine the long-term stability of the coating, the TiO₂-coated filters were subjected to water washing and recirculation, prior to photocatalytic reactions. The TiO₂ thin films were sufficiently robust, with no TiO₂ particles detected in the water after washing. The TiO₂-coated ceramic surface was rougher than the bare ceramic surface, which caused minimal changes in morphology. The rougher surface provided a larger surface area for photocatalytic reaction.

EDXS was used to determine the chemical composition of the TiO₂ thin film on the ceramic foam. Signals corresponding to Ti were detected in the two selected measurement areas (Fig. 4(e) and 4(f)). This indicated that the TiO₂ thin films were deposited on the ceramic foam air filters. The spatial distributions of elements in the TiO₂ thin film on the ceramic foam were determined by elemental mapping. Fig. 5(a)-5(f) show bright-field images with O, Al, P, Ti, and Si present in the same region of the sample. The TiO₂-coated ceramic foam mainly consisted of homogenously distributed Ti, Al, P, and O.

Fig. 5. SEM image (a) of the TiO2-coated ceramic foam air filter and (b-f) elemental O, Al, P, Ti, and Si mapping of the image shown in (a), respectively.

XRD was used to investigate the phase composition of the TiO₂ thin film on the ceramic foam. Fig. 6 and Fig. 7 show XRD patterns of the bare and TiO₂-coated ceramic foam filters, respectively. The XRD patterns’ most intense peak was that of anatase TiO₂ at 25.4°, which overlapped with strong peaks of the ceramic substrate. The XRD peak intensities for the TiO₂ on the ceramic foam were relatively weak compared with those from the ceramic substrate (Fig. 6). Thus, only peaks of the ceramic substrate were observed in the XRD patterns (Fig. 6). To determine the crystal phase of the deposited TiO₂, TiO₂ powders were prepared from a similar deposition solution but without coating on the ceramic foam substrate. Fig. 6(3) shows the XRD pattern of the TiO₂ powder sample, which was consistent with anatase. Therefore, the mesoporous TiO₂ thin film on the ceramic foam (Fig. 4(d)) consisted of many small anatase TiO₂ crystallites.

Fig. 6. XRD patterns of the ceramic foam filter substrate (1), TiO2-coated ceramic foam filter (2), TiO2 powders prepared from similar deposition solution for coating (3). “A” represents the most intense peak of the anatase phase. The angle of incidence was fixed at 0.5°, and 2θ was scanned within the range of 20° to 60°.

Fig. 7. Pore size distribution curve of TiO2 powders prepared from the TiO2 solutions. Inset: the corresponding nitrogen adsorption-desorption isothermals.

Fig. 8. XPS survey spectra of the uncoated (1) and TiO2-coated ceramic foam air filters (2).

The BET surface areas of the mesoporous TiO₂ thin films on the ceramic foam substrates could not be measured directly by nitrogen sorption, because the amount of the thin film was too small. Instead, BET surface areas were measured on powder samples prepared by the same procedure as the thin films. Fig. 7 shows the pore size distribution curves calculated from desorption branches of the nitrogen isotherms, by the BJH method using the Halsey equation. Both samples exhibited narrow pore size distributions. The inset shows the corresponding nitrogen isotherm of the TiO₂. The sample exhibited a type-IV isotherm with a sharp decline in the desorption curve, typical of mesoporous solids. Such strong hysteresis is believed to relate to capillary condensation associated with the mesopores. These mesoporous structures allow rapid diffusion of reactants and products, which enhances the rate of photocatalysis.

The chemical composition of the TiO₂ thin film on the ceramic foam filter was determined using XPS. Fig. 9 shows the XPS survey spectra of the TiO₂ thin film on the ceramic foam filter. The film contained Ti, O, C, Al, P, Si and Na. Peaks for the Ti 2p state were observed at a binding energy (E_b) of 458.6 eV, those for O 1s at E_b = 531.9 eV, Na 1s at E_b = 1071.8 eV, P 2p at E_b = 134.5 eV, Si 2p at E_b = 102.6 eV, and Al 2p at E_b = 74.9 eV. The high-resolution XPS spectra (Fig. 9) showed the Ti 2p_3/2 state at 458.7 eV and the Ti 2p_1/2 state at 464.3.8 eV, consistent with the presence of TiO₂. Table 1 shows the atomic percentages of the bare and TiO₂-coated ceramic foam filters. A significant amount of Ti was detected after TiO₂ coating. The XPS results are consistent with the EDX results, indicating the presence of Ti in the samples after TiO₂ coating.

Fig. 9. High-resolution XPS spectra of Al 2p (a), P 2p (b), Na 1s (c), O 1s (d), Si 2p (e), and Ti 2p (f) regions for the TiO2-coated ceramic foam air filter.

The photocatalytic performances of the TiO₂-coated porous ceramic air filters with different pore densities were evaluated from the degradation of NO under UV irradiation. Fig. 10 shows that the NO concentration sharply decreased when the UV light was switched on for all air filters. A higher pore density (e.g. 30 pores per inch (ppi)) resulted in higher photocatalytic efficiency. The TiO₂-coated porous ceramic filters with 30 and 20 ppi degraded 79% and 76% of NO in a single pass, respectively. The air filter with a lower pore density of 10 ppi degraded 65% of NO in a single pass. This indicated that an optimum pore density was critical for photocatalytic efficiency. A low pollutant concentration was employed in this study. The effect of the adsorption of pollutants and water vapor is more pronounced at the ppb-level concentration of pollutants than at the current ppm-level concentrations. In this study, the NO concentration was maintained at <1000 ppb. The ceramic foam filters with a high pore density had abundant adsorption sites for water vapor and pollutants, which enhanced the photocatalytic efficiency. More than 95% of NO was degraded in a single pass for all air filter samples with different pore densities when two TiO₂-coated foam ceramic filters were used.

Fig. 10. Photocatalytic degradation of NO for the TiO2-coated ceramic foam air filters with different pore densities (pores per inch (ppi)) under UV light irradiation. Dimension of air filter used = 30 cm (L) × 30 cm (W) × 1 cm (H); integrated UV intensity within the range of 310 to 400 nm = 540 ± 10 μW/cm2; RH = 65% ± 5%; Temp. = 22 ± 2°C; initial NO concentration = 1000 ppb.

Table 1 Atomic concentrations (%) of the uncoated and TiO2-coated ceramic foam air filters from XPS analysis. Sample C O Na Al Si P Ti Uncoated air filter 16.9 59.2 4.5 12.9 3.6 3.0 0.0 TiO2-coated air filter 18.0 58.0 1.9 11.6 2.9 6.2 1.5

Table 1 Atomic concentrations (%) of the uncoated and TiO2-coated ceramic foam air filters from XPS analysis.

Figure 11 shows the photocatalytic oxidation of NO and the conversion of NO₂ from NO for the TiO₂-coated ceramic foam air filter with 30 ppi under UV irradiation. The background concentration inside the test chamber was 16 ppb. After the UV light was switched on, NO₂ functioned as the intermediate in the photocatalytic degradation of NO. The amount of generated NO₂ increased with reaction time, reaching a constant concentration of 58 ppb. NO₂ was continuously formed and transformed into HNO₃ during irradiation. NO_x (NO and NO₂) are major pollutants in our environment, and are harmful to human health. Evaluating the photocatalytic activity of the TiO₂-coated filters based on the NO removal rate is therefore insufficient. More attention should be given to the conversion of NO₂ to HNO₃ because the latter can be easily removed. Previous studies ^{[53, 54, 55, 56, 57]} have shown that NO₂ is the intermediate generated in the photocatalytic degradation of NO, which is supported by the following reactions:

Fig. 11. Photocatalytic oxidation of NO and the conversion of NO2 from NO for the TiO2-coated ceramic foam air filter with 30 ppi under UV light irradiation. Dimension of the air filter used = 60 cm (L) × 30 cm (W) × 1 cm (H); integrated UV intensity within the range of 310 to 400 nm = 540 ± 10 μW/cm2; RH = 65% ± 5%; Temp. = 22 ± 2 °C; initial NO concentration = 1000 ppb.

NO₂ functions as the intermediate during the photocatalytic oxidation of NO, and can further react with OH• to form HNO₃ ^{[33, 34]}. The oxidation of NO is reportedly the major process in the photocatalytic removal of NO. NO is directly oxidized to NO₃^- by some oxidative species generated under solar excitation, such as photogenerated holes and •OH and •O₂^- radicals ^{[58, 59, 60, 61, 62]}:

Thus, NO₂ generated from the photodegradation of NO was adsorbed on the ceramic foam air filters because of their large adsorption capacities. Adsorbed NO₂ was photodegraded to HNO₃, which reduced the amount of NO₂ exiting the system. This result further confirmed that the TiO₂ thin film coated on the ceramic foam filters efficiently removed NO and NO₂.

Repeatability and long-life photocatalytic activity are important for photocatalysis, and in minimizing the frequency of air filter replacements. Fig. 12 shows the effect of the reaction cycle on the photocatalytic activity of the TiO₂-coated ceramic filters, for the photocatalytic degradation of NO with an initial concentration of 400 ppb. After sorption equilibrium, the UV light was turned on for more than 1 h, and then turned off for more than 1 h, to recover sorption equilibrium. This process was repeated three times to evaluate the effect of increasing the reaction cycle on the photocatalytic activity of the TiO₂-coated porous ceramic filters. The NO concentration rapidly decreased to a very low level when the UV lamp was switched on. The NO concentration gradually decreased with prolonged UV irradiation time. The air filter samples with different pore densities degraded 92.5%‒97.5% of NO. The photocatalytic oxidation of NO was similar between reaction cycles, with previous cycles having no evident effect on the photocatalysis of the TiO₂-coated filters.

Fig. 12. Photocatalytic degradation of NO in different repeated cycles for the TiO2-coated ceramic foam air filters with different pore densities (pores per inch (ppi)) under UV light irradiation. Dimension of the air filter used = 30 cm (L) × 30 cm (W) × 1 cm (H); integrated UV intensity within the range of 310 to 400 nm = 540 ± 10 μW/cm2; RH = 65% ± 5%; Temp. = 22 ± 2 °C; initial NO concentration = 400 ppb.

NO was photocatalytically degraded to NO₂, and then further oxidized to HNO₃ upon UV irradiation. Fig. 11 shows that the concentration of the NO₂ intermediate in the outlet gas was maintained at a low level. Thus, most of the NO and NO₂ intermediates were photocatalytically oxidized to HNO₃, which could potentially have progressively adsorbed at TiO₂ active sites. This competitive adsorption between HNO₃ and NO could have decreased the rate of photocatalytic oxidation of NO to NO₂, and thus decreased the photocatalytic activity of the TiO₂ thin film. However, no such decrease in photocatalytic activity was observed based on the above reaction cycle testing. HNO₃ adsorbed at active sites transferred to non-active sites of the TiO₂ coating, or to the surface of the ceramic foam substrate, so active sites on the TiO₂ surface were recovered. Thus, the NO₂ intermediate did not accumulate at TiO₂ active sites, and did not deactivate the photocatalyst.

To further evaluate the effect of the 3D porous foamed structure of the filter on the photocatalytic efficiency of TiO₂ coating, the TiO₂ thin film was prepared on flat ceramic tiles using similar deposition conditions. Table 2 compares the photocatalytic activities for the degradation of NO between the TiO₂ thin film deposited on the porous ceramic foam filters, and that coated on the flat ceramic tiles. The TiO₂ thin film coated on the 3D porous ceramic foam exhibited a higher photocatalytic activity. This confirmed that the photocatalytic oxidation of NO diffused into the ceramic foam, because of its large surface area. The photocatalytic oxidation only occurred on the surface of the TiO₂ thin film on the flat ceramic tiles. No further space was available for NO diffusion, which decreased the photocatalytic efficiency. Another factor affecting photocatalytic performance is the nature of the UV light. Light naturally travels from its source (UV lamp) with some degree of divergence, ensuring that the entire top surface of the air filters was irradiated. The UV light intensity was lower within the 3D porous ceramic foam filters, depending on the penetration depth and scattering. More UV light may have penetrated into lower pore density filters. The photocatalytic activity of the TiO₂-coated 20 ppi porous ceramic filter was similar to or even higher than that of the 30 ppi filter. Further studies are needed to elucidate this phenomenon.

Table 2 Photocatalytic degradation of NO for the TiO2-coated ceramic foam air filters with different pore densities. Sample Initial NO conc. (ppb) Degradation of NO in single pass (%) TiO2-coated ceramic filter-10 ppi 400 92.5 TiO2-coated ceramic filter-20 ppi 400 97.5 TiO2-coated ceramic filter-30ppi 400 95 Pure ceramic filter-10 ppi 1000 0 Pure ceramic filter-20 ppi 1000 0 Pure ceramic filter-30 ppi 1000 0 TiO 2-coated ceramic filter-10 ppi 1000 65 TiO2-coated ceramic filter-20 ppi 1000 76 TiO 2-coated ceramic filter-30 ppi 1000 79 2 × TiO 2-coated ceramic filter-10 ppi 1000 97 2 × TiO 2-coated ceramic filter-20 ppi 1000 98 2 × TiO 2-coated ceramic filter-30 ppi 1000 95 TiO 2-coated ceramic tile 400 72.5 TiO 2-coated ceramic tile 1000 43 The initial NO concentrations = 400 or 1000 ppb. Dimensions of the air filters and tiles used under UV light irradiation = 30 cm (L) × 30 cm (W) × 1 cm (H) or 60 cm (L) × 30 cm (W) × 1 cm (H). Integrated UV intensity within the range of 310 to 400 nm = 540 ± 10 μW/cm 2; RH = 65% ± 5%; Temp. = 22 ± 2 °C.

Table 2 Photocatalytic degradation of NO for the TiO2-coated ceramic foam air filters with different pore densities.

Mesoporous TiO₂ thin films were formed on 3D porous ceramic air filters with a high surface area, via the reverse micellar method. The photocatalytic degradation of NO was evaluated on these mesoporous TiO₂-coated porous ceramic air filters. The efficient photocatalytic degradation of NO resulted from the large effective surface area and high photocatalytic activity of the mesoporous TiO₂ thin film coating. More than 92.5% of NO at 400 ppb was degraded in a single pass for the TiO₂-coated ceramic foam filters with different pore densities. The 3D porous structure of the ceramic filters enhanced flow turbulence and mixing, providing sufficient contact between the reactant gas and TiO₂ surface. A higher pore density of the ceramic foam filter resulted in a higher photocatalytic rate. The TiO₂-coated 3D porous ceramic filters exhibited a higher photocatalytic degradation rate of NO than that of TiO₂-coated flat ceramic tiles. The 3D porous ceramic filters maintained a consistently high degradation rate between reaction cycles, indicating that the photocatalyst did not become deactivated.

This research is financially supported by the Research Grant of the Early Career Scheme (ECS 809813) from the Research Grant Council, Hong Kong SAR Government, the Dean’s Research Fund-Early Career Researchers (04022), the Research Equipment Grant (REG-2), and the Internal Research Grant (R3429) from the Hong Kong Institute of Education. This study is also supported by the grants from the Research Grants Council of the Hong Kong Special Administrative Region, China (PolyU 5204/07E) and the Hong Kong Polytechnic University (GYX75).