Abstract

TiO2 foams have been prepared by a simple mechanical stirring method. Short-chain amphiphilic molecules have been used to stabilize colloidal suspensions of TiO2 nanoparticles. TiO2 foams were characterized by X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), UV-vis absorption spectroscopy, and scanning electron microscopy (SEM). The photoassisted oxidation of NO in the gas phase according to ISO 22197-1 has been used to compare the photoactivity of the newly prepared TiO2 foams to that of the original powders. The results showed that the photoactivity is increased up to about 135%. Foam structures seem to be a good means of improving the photoactivity of semiconductor materials and can readily be used for applications such as air purification devices.

1. Introduction

Heterogeneous photocatalysis has been used as an effective technique for the remediation of chemical wastes. Titanium dioxide (TiO2) is one of the most frequently investigated heterogeneous semiconductor photocatalysts and has been shown to be a relatively cheap and effective material to decompose various kinds of organic and inorganic wastes in both, the liquid and the gas phase [14]. As one of the most promising applications, the decomposition of nitrogen oxides (NOx) in ambient air has been studied intensively. Nitrogen oxides (mainly NO and NO2), which are emitted from sources such as automobiles and boilers, have become a serious environmental problem in urban areas and can also cause ozone depletion, photochemical smog, and acid deposition. Substantial efforts have been undertaken to develop methods to reduce the concentration of NO and of NO2. In particular, the photoassisted oxidation of NOx to nitric acid (HNO3) using TiO2 seems to be an effective, economical, and energy saving process for the treatment of diluted NOx [510].

The photocatalytic properties of TiO2 materials do not just depend on the chemical composition but also on the geometrical microstructure [11, 12]. During the last few decades, there have been numerous reports focusing on methods to enhance the photoactivity of TiO2 materials [13, 14]. Several research groups have been working on the doping of TiO2 with noble metals (Au, Pd, Ag, and Pt) or nonmetal atoms (N, F, S, C). All these investigations have developed our knowledge concerning TiO2-based photocatalysts and have generated very interesting and important results in this research area. However, in general, the TiO2-based photocatalysts suffer from rather low quantum efficiencies because of poor photoabsorption efficiencies and weak molecular transport capabilities [5, 1522].

Another strategy to enhance the photoactivity of TiO2 materials is the creation of new geometrical microstructures. It has, for example, been reported that the photoactivity of TiO2 with structural, that is, ordered porosity is higher than that of nanoparticular TiO2 with only interparticular nanovoids. Materials with structural porosity have high surface areas, and all pores are well interconnected. The macro- and mesoporosity improves the adsorption and diffusion of reactants and products. The existence of cavities furthermore increases the light-capturing efficiency because of their strong light-scattering properties [21, 2327].

Titanium dioxide foams have shown to have a higher photoactivity as the powder one by Zhao et al. [24]. However, it is not clear how the foam structure affected the photoactivity. One reason could be changes in the surface properties of TiO2. Further investigation is necessary to confirm this theory. Photocatalytic deposition velocity is known to be an input value that predicts the possible effect of photocatalytically active surfaces on air pollution in urban areas.

In this article, it is reported that TiO2 foams have been prepared by a simple mechanical stirring method employing short-chain amphiphilic molecules to surface-lyphobize commercially available TiO2 nanoparticles. The aims of this work were to study the photoactivity and the changes in the surface properties of TiO2 foams prepared in this way and compare it to that of the starting TiO2 powder. In 2007, ISO published a standard test to evaluate the photoassisted oxidation efficiency of the air-cleaning products. The photoassisted degradation of NO in the gas phase according to ISO 22197-1 was used as the test reaction to compare the photoactivity of the TiO2 foams to that of the original powders. Photocatalytic deposition velocities were determined to describe the changes in the surface properties. The morphology and the structural properties of TiO2 foams have also been studied.

2. Experimental

2.1. Materials

Commercial titanium dioxide Evonik AEROXIDE® P25 (30% rutile, 70% anatase, 20–30 nm) has been used as received. Hexylic acid (hexanoic acid) and hexylamine from Sigma-Aldrich have been used as short-chain amphiphilic molecules. Other chemicals used in the experiments were deionized water and sodium hydroxide (Sigma-Aldrich).

2.2. Foam Preparation

Titanium dioxide foams were prepared according to the method described by Zhao et al. [24]. In short, 2 g of P25 was added stepwise to water under vigorous stirring for 20 minutes. The pH value of the suspension was around 3.6. The suspension was homogenized for about 30 min using an Ultrasonic Vibra-Cell. Meanwhile, a solution containing short-chain amphiphilic molecules (0.3 g hexylic acid in water, the ratio of TiO2 to amphiphilic molecules was adapted from [24]), was prepared. This solution was then added dropwise to the TiO2 suspension under vigorous stirring for 30 min.

The pH was adjusted to pH 3.5 with 0.1 M NaOH if necessary (in case of using hexylic acid as amphiphilic agent). Finally, the wet TiO2 foam was produced and this wet foam was air dried for one day at room temperature. Figure 1 shows photographs of freshly prepared wet TiO2 foam and of a dry one.

2.3. Characterization

The crystal structures of the samples were determined by X-ray powder diffraction at room temperature using Cu Kα radiation on a Bruker D8 Advance diffractometer in the 2θ range of 10–80°. The particle size distributions were calculated from the peak broadening of the XRD patterns using the Scherrer equation [28]. X-ray photoelectron spectroscopy (XPS) measurements were carried out using a Thermo Fisher ESCALAB 250Xi equipped with monochromatic Al-Kα X-ray source (1486.6 eV). The binding energies (BE) were referenced to the adventitious carbon contamination C 1s peak (284.8 eV). High-resolution XPS spectra of Ti 2p, O 1s, and C 1s were recorded with 10 eV pass energy, and 0.02 eV step size with 5 spectra was averaged.

A scanning electron microscope (SEM) (JEOL JSM-6700F field-emission) was used to characterize the morphology of the TiO2 foams, and the Varian Cary 4000 UV-vis spectrophotometer was used to measure the diffuse reflectance spectra of the samples. Band gap energies were calculated by using Tauc plots resulting from the Kubelka-Munk transformation of the respective diffuse reflectance spectra. The nitrogen sorption isotherms were measured at 77 K using a Quantachrome Autosorb 3. The foam samples were outgassed in vacuum at 25°C and P25-TiO2 at 100°C for 24 hours prior the sorption measurement. The sorption data were analyzed with the quantachrome software ASiQwin 2.0.

2.4. Photoassisted Oxidation of NO

The photoassisted oxidation of NO was measured according to ISO 22197-1 [29]. The experimental details were described previously [9, 10, 30]. Briefly, the test equipment included the test gas supply, a humidifier, three mass flow controllers (Brooks Instrument), a photoreactor, a light source (Philips, Cleo Compact, lmax = 355 nm, 15 W), and a chemiluminescent NO–NOx analyzer (Horiba APNA 360). The concentration of the probe gas was followed by a NO/NO2 analyzer in the dark until equilibrium was reached. The probe gas mixture had a concentration of 1 ppmv NO (Linde, 50 ppmv NO in N2), a relative humidity of 50%, and a laminar volume flow V̇ of 5 × 10−5 m3/s. For a single degradation test, TiO2 foam or the P25 sample was placed into the photoreactor with a geometric surface area of 18.49 cm2 and covered with a UV(A) transparent borosilicate glass. The NO concentration was adjusted in the dark via a bypass mode followed by a dark adsorption of the pollutant on the sample surface by switching from bypass to reaction mode. After equilibration, the degradation of NO was followed under irradiation (light intensity 10 W/m2) for 2 h until a steady state was reached. Subsequently, the system was kept in the dark until the initial concentration of NO was achieved again. For the determination of the photocatalytic deposition velocity use, a photoreactor with a geometric surface area of 39.6 cm2 and the amount of the probe gas (NO) have been changed during the photoassisted oxidation. The NO concentration varied between 0.1 ppmv and 1.0 ppmv. The flow rate of the probe gas (5 × 10−5 m3/s), the temperature (25°C), and the relative humidity (50%) inside the photoreactor were kept constant.

3. Results and Discussions

3.1. TiO2 Foam Characterization

The scanning electron microscopy (SEM) images of the TiO2 foam show a porous, spongy, and rough morphology. The photograph and the low magnification image of the foam reveal macroporosity (Figure 2). The pore size distribution of the foams depends on the type of amphiphilic agent that has been used. For example, TiO2-foam-1 (with hexylic acid) has an average pore size of approx. 0.24 ± 0.09 cm and TiO2-foam-2 (with hexylamine) has an average pore size of approx. 0.11 ± 0.04 cm. These large pores with extended networks provide many canals for diffusion of reactants. With increasing SEM magnification, the primary TiO2 nanoparticles can be imaged (particle size distribution: 20–30 nm). Supplementary 1 presents the SEM photographs of TiO2-P25, TiO2-foam-1, and TiO2-foam-2.

The UV-vis diffuse reflectance spectra (Supplementary Material, Fig. S2) from P25 and from the TiO2-foam-1 have been measured, and the energy band gaps of the materials were obtained using the Kubelka-Munk function. The band gap energy of the materials was not changed (Eg ~ 3.2 eV) indicating that the semiconductor properties of the TiO2 foam are same as those of the TiO2 powder.

BET surface areas have been determined from the N2 adsorption/desorption measurements (see Supplementary Material, Fig. S3). There is no big difference between BET surface areas of the P25 and to those of TiO2-foam-2 (with hexylamine). P25 shows a surface area of 58.9 ± 1.8 m2·g−1 whereas TiO2-foam-1 (with hexylic acid) has a surface area of 51.7 ± 1.6 m2·g−1 and TiO2-foam-2 (with hexylamine) has a surface area of 58.1 ± 2.5 m2·g−1. Based on the experimental data, we recognize that the nitrogen sorption is not a suitable method to determine the pore size distribution of our foams. It is obvious that the TiO2 foams have large pore size which this method could not determine.

Figure 3 shows the XRD pattern of the TiO2 foam. This XRD pattern is, as expected, consistent with the anatase and rutile crystal phases of TiO2 (JCPDS file number anatase: 00-0210-1272; rutile: 01-070-7347). The three characteristic sharp peaks at 2θ = 25.2°, 37.7°, and 48.0° correspond to the (101), (044), and (200) planes of the anatase phase, respectively, and the peaks at 2θ = 27.4°, 54.3°, and 36.0° fit to the (110), (211), and (101) planes of the rutile phase, respectively. The average particle sizes of the TiO2 nanoparticles in the foam were calculated using the Scherrer equation yielding approx. 17 nm for anatase and 20 nm for rutile, which is in agreement with the SEM observations. Supplementary 4 presents a comparison between the XRD pattern of TiO2-foam-1 with original used P25 (Fig. S4) and the XRD pattern of all three samples (TiO2-P25, TiO2-foam-1, and TiO2-foam-2; Fig. S5). All three samples have the same crystal structure, which is a mixture of anatase and rutile phases.

The surface composition and chemical states of the TiO2-foam-1 and the starting TiO2 powder (P25) were investigated by XPS. The survey spectra for both samples show exclusively the presence of C, O, and Ti (see Figure 4(a)). The Ti 2p photoelectron line (Figure 4(b)) consists of the Ti 2p3/2 and Ti 2p1/2 doublets at binding energies of 458.9 (459 eV) and 464.6 (464.7 eV), respectively. These values are closely related to the reported Ti 2p lines ranging 458.5–459.5 eV and 464.2–464.7 eV, respectively [31, 32].

The peak separation (~5.7 eV) and the strong satellite peak 14 eV above the main line indicate the Ti4 oxidation state. The main O 1s line (Figure 4(c)) is located at 530.1 (530.2 eV) assigned to the lattice oxygen with additional broad peak centered at 531.1 (530.9 eV) related to low coordinated surface O atoms (with low electron density than the lattice O2− ions), respectively (PCCP 2000, 2, 1319–1324). The Ti : O ratios were found to be 0.48 and 0.52 for the TiO2 foam and the P25, respectively.

3.2. Photoactivity of TiO2 Foams

The photoassisted degradation of NO in the gas phase was used to compare the photoactivity of the TiO2 foams to that of the original powders (P25). The mechanism and kinetics of this reaction have been studied intensively before [9, 10]. It has been reported that using Degussa P25 as a catalyst, the determination of the reaction rate by the Langmuir–Hinshelwood model has some limitations. Wang et al. reported about the effect of variations of the inlet NO concentrations. Their observations showed that the reaction is first-order at low concentrations and zero-order at high concentrations [31, 32]. Studies conducted by Dillert et al. have shown that the NO oxidation at TiO2 surfaces also depends on the light intensity. According to their analysis, the rates of the photocatalytic NO oxidation on UV(A)-irradiated TiO2 samples at varying nitrogen(II) oxide concentrations and photon fluxes and possibly at varying humidity can be predicted by a mathematical model [10].

Figure 5 shows the concentration changes of NO, NO2, and NOx as a function of UV illumination time during the photoassisted oxidation of NO with employing AEROXIDE P25 and TiO2 foam, respectively, under the same conditions. The typical experimental data for NO oxidation can be divided into four zones. Zone A is characterized by the adjustment of the NO concentration (1 ppm) before UV light illumination. Zone B is characterized by the adsorption of NO in the dark followed by the saturation of the surface of the photocatalyst sample. Once there is no further change in the NO concentration, the light source can be switched on to start the photocatalytic reaction (Zone C). After 2 h of irradiation with UV light, the light source is switched off (Zone D).

The photoassisted oxidation of NO is assumed to be a surface reaction mediated between NO and photogenerated hydroxyl radicals. And its mechanism has been already investigated. The three major steps during the photoassisted oxidation are oxidation of NO to HNO2, oxidation of HNO2 to NO2 (in the transient state), and finally oxidation of NO2 to HNO3. If the catalyst is saturated with HNO3, a steady state is attained and the oxidation reaction can go only as far as to NO2 [2, 9, 10, 33, 34].

Considering the reaction dynamics during the entire photocatalytic process, the photocatalytic efficiency is determined by the number of photogenerated charge carriers which can avoid the recombination reaction. The photonic efficiency (ζ) of the NO degradation is calculated from the degradation rate of NO and the incident photon flux according to the following equation [33]:

The incident photon flux (J0) is calculated based on (2) with its value being (UV-A light intensity 1 mW·cm−2, average illumination wavelength , irradiated surface area 18.49 cm2, NA Avogadro number, h Planck constant, and c light velocity). where I is the UV-A light intensity (1 mW·cm−2), Ar is the illuminated surface area (18.49 cm2), λ is the average illumination wavelength (350 nm), NA denotes Avogadro number; h denotes Planck constant; and c denotes light velocity.

The degradation rate of NO is calculated using the following equation (where qV is the volumetric flow rate):

The results are summarized and tabulated in Table 1. In this table, the results are presented as degraded concentration (Cgas in ppm), calculated degradation rate (∆nNO/∆t in mol·s−1), and as photonic efficiency (ζ in %). The photonic efficiency (ζ %) for the complete degradation of the incoming NO gas under ISO 22197-1 conditions for a sample with a surface area of 18.49 cm2 is approx. 4% (for a sample with a surface area of 50 cm2, it is 1.43%).

From the results presented in Table 1, the following conclusions can be drawn: the TiO2-foam-1 (using hexylic acid as amphiphilic molecule) adsorbs more NO than P25 powder and produces more NO2 and NOx (HNO3). This could be an indication of the activity of TiO2 foams under the UV illumination. And also it shows that the foam structures are photocatalytically more active than the original powders.

Engel et al. [30] published an experimental method for the determination of the photocatalytic deposition velocity. To gain more information about our samples and their differences, we used this method. For this investigation, we prepared two different TiO2 foams (using different amphiphilic molecule) and change the amount of the probe gas (NO) during the photoassisted oxidation. The following equation has been used to calculate photocatalytic deposition velocity:

This equation is a linear equation of the two variables and having the slope and the intercept with the ordinate Ki. With the calculation of the slope, the photocatalytic deposition velocity will be determined.

Figure 6 presents the change of the NO concentration at the reactor outlet during irradiation of AEROXIDE P25, TiO2-foam-1(with hexylic acid), and TiO2-foam-2 (with hexylamine) with varying NO inlet concentrations. All other experimental parameters, that is, temperature, relative humidity, and UV (A) photon flux, were kept constant during these experimental runs. With the value of illumination surface area Ar (3.96·10−3 m2), the volumetric flow rate qV (3.0 L min−1), and the measured values of the NO inlet and outlet concentration, the values of the two variables of (4) have been calculated and plotted for each individual experimental run. These plots clearly show that the TiO2-foam-2 (with hexylamine) has the more active surface for the NO oxidation and both foams are photocatalytically more active than the original powders.

The photoactivity depends on various factors such as the employed catalyst, the light collection efficiency, and the rate of molecular diffusion [7, 33, 35]. The activities of catalysts with the same chemical compositions (P25) have been compared (Table 1 and Figure 5). The light intensity for all experiments was the same although the light collection by the catalyst could be influenced by its structure. The interesting point from our results is that due to the macroporous/mesoporous structure of the foam, the photocatalyst did not own large surface area. The N2 sorption results could not show big difference between the surface area of P25 and that of the foams. Although there is no big difference between the surface areas, the foam structures show more photoactivity.

TiO2 foam shows high photoactivity due to the structure of macroporous/mesoporous channels that provide fast intraparticle molecular transfers, which improves the light harvesting and the adsorption of reactant molecules. This macroporous/mesoporous structure also provides proper interfaces for a facile interparticle charge transfer while the reactants can freely diffuse through the pores. At the same time, the light absorption and the energy transfer through the foam structure into the inner surface of the macroporous/mesoporous TiO2 are apparently as high as expected.

It has been shown that mass transfer and surface reactions control heterogeneous catalytic reactions [10, 36]. Yang et al. [35] reported that the mass transfer coefficient is affected by the flow velocity. Beyond a certain flow velocity, the photocatalytic reaction process was founded to be independent of the mass transfer and was only controlled by the surface reactions. In this work, the value of the flow velocity is rather high and identical for all reactions. Therefore, the photocatalytic process is apparently independent of any mass transfer limitations and will only be influenced by surface properties.

The surface properties of the photocatalyst play an important role as heterogeneous photocatalytic reactions usually take place at the interface of solid/liquid or solid/gas phases. The large surface of the foam structure is apparently very effective due to the fact that the photoassisted oxidation of NO in the gas phase requires the adsorption of the reactants at the TiO2 surface. The amount of adsorbed substances increases in the TiO2 foam and thus enhancing the photocatalytic process.

4. Conclusions

Titanium dioxide foam was prepared by a simple mechanical stirring method. The obtained TiO2 foam has been characterized using X-ray powder diffraction and SEM. The crystal structure of the TiO2 foam remains unchanged as compared with the initial TiO2 particles. The SEM investigations showed very rough and porous surface structures of the TiO2 foam. These macroporous/mesoporous foams were used to measure the photoassisted degradation of NO according to ISO 22197-1. The photocatalytic properties of the TiO2 foam, compared to those of the original photocatalyst powder, improved significantly and increased by about 135%. The photocatalytic properties of TiO2 materials do not only depend on their chemical composition but also on their geometrical microstructure. The porous structure is an attractive way to achieve high photocatalytic activities.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Acknowledgments

The authors thank F. Steinbach (PCI) Leibniz Universität Hannover for his support during the SEM measurements and M. Jahns and A. Mohmeyer (ACI) Leibniz Universität Hannover for their support during the N2 sorption measurements. The authors also thank Dr. D. H. Taffa from Carl von Ossietzky Universität Oldenburg for the XPS measurements. Maryam Jami would like to thank Hannover School for Nanotechnology for her doctoral scholarship.

Supplementary Materials

Supplementary 1. Fig. S1: (a) SEM photographs of TiO2-P25 (Evonik AEROXIDE P25) with different magnifications. (b) SEM photographs of TiO2-foam-1 (hexylic acid as amphiphilic molecule) with different magnifications. (c) SEM photographs of TiO2-foam-2 (hexylamine as amphiphilic molecule) with different magnifications. Supplementary 2. Fig. S2: comparison between the diffuse reflectance spectra of TiO2-P25 and prepared TiO2 foam. Inset shows the modified Kubelka-Munk function versus the photon energy of these two samples. Supplementary 3. Fig. S3: left: nitrogen adsorption-desorption isotherm of TiO2-foam-1 (hexylic acid as amphiphilic molecule), TiO2-foam-2 (hexylamine as amphiphilic molecule), and TiO2-P25; right: the BJH graphs indicated corresponding pore size distributions of samples. Supplementary 4. Fig. S4: comparison between the XRD pattern of TiO2 foam (with hexylic acid) and commercial used TiO2 (Evonik AEROXIDE P25). Fig. S5: XRD patterns of TiO2-P25, TiO2-foam-1, and TiO2-foam-2, the respective reflection peaks of anatase (A) and rutile (R) phases are labeled with their Miller indices. (Supplementary Materials)