Enhancement of the photocatalytic activity of a TiO₂/carbon aerogel based on a hydrophilic secondary pore structure

Abstract

Improving the separation and utilization of electrons and holes in a photocatalytic process is a guarantee for high photocatalytic efficiency. We report a strategy to enhance the photocatalytic performance based on fabrication of a hydrophilic secondary pore structure by incorporating TiO₂ into a porous carbon aerogel (CA) with a 9.3 nm pore diameter, where TiO₂ resides on both the inner and outer surfaces of CA as evidenced by N₂ sorption isotherms and transmission electron microscopy. In such a structure, the spatial separation efficiency of photoelectrons and photoholes is supposed to get enhanced with interface electrons transferring into the inner surface of the pores via conductive CA. As a result, HO˙ formation can be promoted in the confined inter surface of the hydrophilic secondary channel through O₂ reduction with the participation of photoelectrons and H₂O. And the remaining photoholes on the outer surface can oxidize water to generate HO˙ as well. In contrast, TiO₂ is mainly dispersed on the outer surface of CA as small pore diameters of 3.4 and 4.3 nm; as a result, only uncombined photoholes on the outer surface contribute to HO˙ generation via the water oxidation route. In line with this understanding, TiO₂/CA (9.3 nm) shows the largest amount of HO˙ and thereby the highest efficiency of dimethyl phthalate degradation, as respectively evidenced by the electron paramagnetic resonance spectroscopy and the photocatalytic degradation test. These findings unveil the contribution of the surface/interface synergy effect on the separation and utilization of electrons and holes in photocatalytic process, and provide a potential strategy to enhance the photocatalytic performance.

---

Introduction

As an advanced oxidation process, photocatalytic degradation mediated via TiO₂ is an efficient method to eliminate organic pollutants from water. It has drawn much attention in the field of removing phthalate esters (PAEs) from an aqueous environment, because the end products after photocatalysis on PAEs are carbon dioxide, water, and inorganic mineral ions.^1–7 For TiO₂ photocatalysts, improving the separation and utilization of electrons and holes in photocatalytic processes is highly desired for high photocatalytic efficiency.⁸ Surface modification,⁹ sensitization¹⁰ and transition-metal doping¹¹ are widely used strategies to optimize TiO₂ photocatalytic performance. In addition, assembling TiO₂ with carbon materials is another effective way,¹² as porous carbon support has been found to selectively accumulate reactants into its abundant pores^13–15 and its excellent conductivity serves to accelerate the separation of photogenerated carriers.^16–18 Some studies have suggested that such porous carbon powder is also capable of absorbing light, and the generated photoelectrons can be transferred to the conduction band of TiO₂ for further use.¹⁹ Despite a large number of assemblies of TiO₂ with carbon were reported to data,^17,20–23 few systematic studies are undertaken to reveal the function of pore structure on photocatalytic performance, especially secondary pores with unique spatial features that are formed by encapsulating TiO₂ into carbon pore cavities. It has been reported that the unique pore-confined space can potentially play a vital role in enhancing catalytic activity and selectivity in heterogeneous catalysis.^24–26 Particularly, the CNT-confined TiO₂ exhibited improved visible-light activity as the modification of the electronic structure of TiO₂ induced by the unique confinement inside CNTs.²⁷

Carbon aerogel (CA), possessing unique three-dimensional ultra-porous network with high porosity (80–98%), low electrical resistivity (≤40 mΩ cm), and large surface area (400–1100 cm² g⁻¹),^28,29 is a promising material in the application of processing contaminants.^28,30 In this work, three kinds of mesoporous CA with several nanometer pore diameters³¹ were employed as support to assemble TiO₂/CA-X composite catalysts, where X denotes as the pore diameter of CA. Dimethyl phthalate (DMP), a member of phthalate acid esters (PAEs) detrimental to human being as endocrine disruptors, mutagens, reproductive and developmental toxicants,^7,32,33 was chosen as the target pollutant for evaluating the photocatalytic properties of the TiO₂/CA catalysts. Only TiO₂/CA-9.3 shows the enhanced photocatalytic activity, and TiO₂/CA-3.4 and TiO₂/CA-4.3 even gives the lower activity than bare TiO₂. It was found that the HO˙ formation is correlated with the secondary pore structure of TiO₂/CA-X, and the contribution of surface/interface synergy effect on the separation and utilization of electrons and holes in the photocatalytic performance is discussed.

Results and discussion

Pore structure of TiO₂/CA-X

The pore structure of CA-X before and after assembling TiO₂ are investigated by N₂ adsorption–desorption isotherms, and the results are shown in Fig. 1a and b, respectively. The isotherms of CA-X and TiO₂/CA-X featured IV-type character with specific hysteresis loop at relatively high pressure, implying that both CA-X and TiO₂/CA-X are typical mesoporous materials. Fig. 2 shows the pore distribution of CA-X and TiO₂/CA-X samples calculated by a DFT model on ASiQwin software.^34,35 It can be seen that mesopores dominate the pore structure in both CA-X and TiO₂/CA-X samples with a small proportion of micropores and macropores. The detailed information of the pore structure is summarized in Table 1, e.g. pore volume, pore diameter and BET surface area.

Fig. 1 N₂ adsorption–desorption isotherms of (a) CA-X and (b) TiO₂/CA-X and pure TiO₂.

Fig. 2 Pore distribution of (a) CA-X and (b) TiO₂/CA-X and pure TiO₂.

Table 1 Pore parameters of carbon, TiO₂, and TiO₂/CA

Sample	D^a (nm)	SBET^b (m^2 g^−1)	Vmicro^c + Vmeso^d (cm^3 g^−1)	DMP adsorption amount (mg g^−1)
a Average mesopore size calculated form DFT method.b Brunauer–Emmett–Teller (BET) surface area calculated from adsorption branch of the N2-sorption isotherm.c Micropore (<2 nm) volume calculated from cumulative pore volume of DFT method.d Mesopore (2–50 nm) volume calculated from cumulative pore volume of DFT method.
CA-3.1	3.1	1173	0.88	8.6 ± 0.4
CA-4.3	4.3	954	0.96	7.7 ± 0.5
CA-9.3	9.3	709	1.2	10.6 ± 0.6
TiO2	4.4	84	0.07	0 ± 0.2
TiO2/CA-3.1	3.8	237	0.17	6.4 ± 0.5
TiO2/CA-4.3	4.1	215	0.18	6.2 ± 0.5
TiO2/CA-9.3	6.1	124	0.17	4.4 ± 0.5

---

The average pore diameters of three pristine CA samples are 3.1, 4.3 and 9.3 nm, as shown in Table 1. After assembling TiO₂, the pore diameters of CA-3.1 and CA-4.3 remain almost unchanged, but it decreases from 9.3 to 6.1 nm for CA-9.3. As seen in Table 1, both the volume of micropore (<2 nm) and mesopore (2–50 nm) and the BET surface of three TiO₂/CA-X samples decrease sharply when compared with those of the pristine CA-X samples. The decrease can be attributed to either the blockage of pore channels by TiO₂ dispersed on the carbon surface or the shrinkage of pore channels as TiO₂ was encapsulated into pore channels. The unchanged pore diameter of CA-3.1 and CA-4.3 indicates that TiO₂ was prominently dispersed on the outer surface of CA-3.1 and CA-4.3. While for CA-9.3, the dwindled 6.1 nm pore size suggests that, in addition to being dispersed on the outer surface, a considerable amount of TiO₂ was dispersed into the pore channels of CA-9.3. It is acknowledged that the capillary force is the driving force to infiltrate the precursor solution into the support channels during the preparation of supported catalysts,³⁶ which could account for the formation of secondary pore in TiO₂/CA-9.3. The 3.1 and 4.3 nm pore diameters of CA may be too small to allow the infiltrate of Ti precursor into the pore channel.

Composition and structure of TiO₂/CA-X

The carbon content of TiO₂/CA-X samples was analyzed by thermogravimetry (TG) and differential thermal gravity (DTG) (Fig. S1†). The same loss steps at around 500 °C related to the carbon combustion indicate three samples possess the comparable carbon contents of about 15 wt% (Table 2), that is, the same amount of TiO₂ is dispersed on them. X-ray diffraction was used to determine the crystal structure of pristine TiO₂, TiO₂/CA-X and CA-X, and diffraction patterns are shown in Fig. S2.† A set of diffraction peaks related to anatase (PDF#21-1272) are observed in pristine TiO₂ and TiO₂/CA-X, in comparison, CA-X presents amorphous structure with broad diffraction patterns shown in insert of Fig. S2.† The results indicate that three TiO₂/CA-X samples possess the same TiO₂ crystal structure, irrespective of the pore diameter of CA.

Table 2 Carbon content of TiO₂/CA-X samples

Sample	C content (%)	TiO2 content (%)	C/TiO2 ratio
TiO2/CA-3.1	15.0	85.0	0.18 : 1
TiO2/CA-4.3	14.9	85.1	0.18 : 1
TiO2/CA-9.3	15.1	84.9	0.18 : 1

---

The structure of CA-X before and after assembling TiO₂ was further studied by TEM and HRTEM, and the representative images of these samples are shown in Fig. 3. It can be seen that both CA-4.3 (Fig. 3a) and CA-9.3 (Fig. 3c) exhibit disordered wormhole-like mesopore structures, which are consistent with the previous report.³¹ After assembling TiO₂ on CA-4.3 (Fig. 3b), abundant and spherical domains (marked in the dash line frame) on several nanometer scale with clear lattice fringes of TiO₂ are disorderly dispersed on the surface of CA-4.3. When comparing CA-9.3 (Fig. 3c) vs. TiO₂/CA-9.3 (Fig. 3d), the pore channels of CA-9.3 are partially filled by the assembled TiO₂, deduced from the variation on contrast of the two images. In addition, TiO₂ particles (marked in the dash line frame) with identifiable lattice fringes are observed on the edge of CA-9.3 without serious aggregation. The TEM results further confirm that TiO₂ is dispersed on outer surface of CA-4.3, while on both inner and outer surfaces of CA-9.3, which are consistent with the pore structure analysis (vide supra).

Fig. 3 HRTEM images of (a) CA-4.3, (b) TiO₂/CA-4.3, (c) CA-9.3, and (d) TiO₂/CA-9.3.

Effect of porous structure and wettability on DMP adsorption

The DMP adsorption on TiO₂/CA-X is evaluated by kinetic adsorption curves in Fig. 4, and the one-hour absorption amount is listed in Table 1. CA-3.1 and CA-4.3 present the similar adsorption behaviors (dashed line) with ca. 8 mg g⁻¹ DMP amount upon one-hour absorption, which is less than the amount of 10 mg g⁻¹ on CA-9.3. After assembling TiO₂, DMP adsorption on three TiO₂/CA-X samples (solid lines) decreases in different levels compared with each pristine CA-X. It decreases from 10 to 4 mg g⁻¹ when comparing CA-9.3 with TiO₂/CA-9.3, and from ca. 8 to 6 mg g⁻¹ upon assembling TiO₂ on CA-3.1 or CA-4.3. It was reported the volume of mesopores and micropores determined the adsorption of organic micro pollutants, such as microcystins by activated carbons.^37,38 The total volume of micropores and mesopores for CA-3.1, CA-4.3 and CA-9.3 (listed in Table 1) is about 0.88, 0.96 and 1.2 cm³ g⁻¹, respectively. The volume data well accounts for the similar DMP adsorption capacities of CA-3.1 and CA-4.3 and the highest adsorption amount of CA-9.3. After assembling TiO₂ on CA-X, the total volume of all three samples decreases significantly to the same level of ∼0.17 cm³ g⁻¹, as presented in Table 1. The volume decrease of all three samples are consistent with their adsorption amount decrease, but the comparable 0.17 cm³ g⁻¹ total volume cannot explain their different adsorption capacity, e.g. 4 mg g⁻¹ in TiO₂/CA-9.3 and the 6 mg g⁻¹ in TiO₂/CA-3.1 and TiO₂/CA-4.3.

Fig. 4 The adsorption curves for DMP onto CA-X, TiO₂/CA-X, and pristine TiO₂.

In addition to pore volume, the wettability of adsorbent also plays a vital role on its absorption performance, for example, the hydrophilic TiO₂ possesses the negligible DMP adsorption in Fig. 4. Fig. 5 presents the contact angle results and the droplet images of all samples as well. It can be seen that all CA-X samples are hydrophobic with apparent contact angles over 120°. After loading hydrophilic TiO₂ with a contact angle of less than 10°, all samples change to be hydrophilic accompanying with the decrease of contact angles. Notably, TiO₂/CA-9.3 shows the smallest contact angle ca. 70° in three TiO₂/CA-X samples, which is consistent with its unique pore structure that hydrophilic TiO₂ was dispersed on both inner and outer surface of CA-9.3. The hydrophilic nature of TiO₂/CA-9.3 enables it to be more inclined to adsorb H₂O but not DMP. Thus, TiO₂/CA-9.3 gives the lowest DMP adsorption amount, as shown in Table 1.

Fig. 5 Contact angle and droplet images of CA-X and TiO₂/CA-X.

Photocatalytic degradation of DMP

DMP was chosen as the target pollutant to study the influence of pore structure on the photocatalytic activities of TiO₂/CA-X samples. It should be noted that no DMP degradation occurred when irradiating TiO₂/CA-X photocatalysts with visible light (λ > 420 nm). As previous reports,^39,40 the photodegradation process in our case occurs with UV-light irradiation of photocatalysts and carbon does not sensitize TiO₂, though the visible absorption related to carbon is observed in UV-vis absorption spectra (Fig. S3†).

Fig. 6 shows the dynamic curves of DMP photocatalytic degradation during 3 hours. It can be seen that TiO₂/CA-9.3 gives the highest degradation efficiency of 83%, and TiO₂ shows the secondary degradation efficiency ca. 70%. TiO₂/CA-3.1 and TiO₂/CA-4.3 respectively exhibit the third and fourth degradation efficiency of 55% and 44%, which are followed by the mechanical mixture of TiO₂ and CA-9.3 with 25% efficiency.

Fig. 6 Dynamic curves of photocatalytic degradation DMP with TiO₂/CA-X, TiO₂ and TiO₂ + CA-9.3.

The intermediates in DMP degradation process are identified from positive modes of MS (Fig. S4†). The signal at m/z = 195 (+H) which belongs to DMP has been strongly weakened and two major intermediates were formed according to the signals in the mass spectra. The signals at m/z = 149 (+H) and 171 (+Na) might belong to o-phthalic anhydride. The signal at m/z = 167 (+H) might be phthalic acid. The similar intermediates were observed by Wang et al.³

Photocatalytic activity is acknowledged to rely on the adsorption capacity of organic molecules upon photocatalyst, and high adsorption capacity is favorable for the photocatalytic degradation due to the accumulation of organic molecules and the intermediates, if any.⁴¹ However, in our case, the high adsorption amount of DMP in TiO₂/CA-3.1 and TiO₂/CA-4.3 did not result in high degradation efficiency. TiO₂/CA-9.3, which has lowest adsorption amount of DMP, presents the highest degradation efficiency. That is, the DMP accumulation in TiO₂/CA-X relying on the adsorption capacity is unessential for its photodegradation in this case.

In consideration of the similar absorption features of TiO₂/CA-X in UV-vis absorption spectra (Fig. S3†), the amounts of photogenerated carriers in these samples are comparable, which cannot account for such a big difference in photodegradation efficiency (see Fig. 6) as well. The HO˙ was further studied to address the difference in DMP degradation efficiency. Fig. 7 gives the EPR spectra of DMPO-OH adduct produced by in situ illuminating aqueous solutions respectively containing TiO₂, TiO₂/CA-9.3 and TiO₂/CA-3.1. The control experiment result of DMPO aqueous solution is also shown in Fig. 7. It can be seen that TiO₂/CA-9.3 exhibits the largest amount of HO˙, TiO₂ gives the secondary HO˙ amount, and TiO₂/CA-3.1 shows the lowest HO˙ amount. This order of HO˙ amount is consistent with the DMP degradation sequence in Fig. 6. It was reported that water treatment processes was highly dependent on hydroxyl radical and it is critical to maximize hydroxyl radical generation during TiO₂ photocatalysis.⁴ In light of above MS results, the major intermediates should be formed by the attack of hydroxyl radicals on the alkyl chains of DMP. Therefore, we focus on the formation of HO˙ to correlate the structures of TiO₂/CA-X with their photocatalytic properties hereinafter.

Fig. 7 EPR spectra of DMPO-OH adduct in aqueous solution respectively containing TiO₂, TiO₂/CA-9.3 and TiO₂/CA-3.1 illuminating with Xe lamp.

The mechanism of photocatalytic degradation

A typical photocatalytic process begins with the light absorption to produce electrons and holes (eqn (1)). The photogenerated holes are consumed by the oxidization of H₂O to produce active HO˙ (eqn (5));⁴² meanwhile, the photoelectrons are consumed by O₂ reduction to ˙O²⁻ (eqn (2)), followed by the generation (eqn (3)) and decomposition (eqn (4)) of H₂O₂, in which process active HO˙ is also generated. It is noted that active HO˙ with strong oxidation ability leads to the fast degradation of organic molecules (eqn (6)).⁴³

TiO_2(surface) + hν → e_CB⁻ + h_VB⁺ (1)

e_CB⁻ + O₂ → ˙O₂⁻ (2)

˙O₂⁻ + H₂O → H₂O₂ (3)

H₂O₂ → HO˙ + OH⁻ (4)

H₂O + h_VB⁺ → HO˙ + H⁺ (5)

HO˙ + C₁₀H₁₀O₄ → H₂O + CO₂ (6)

In above processes, there are two ways for HO˙ formation: H₂O oxidation in eqn (5) and O₂ reduction with the participation of H₂O in eqn (2)–(4). Most important of all, in both ways, H₂O molecule is indispensable for HO˙ formation. Thus, the more hydrophilic surface in TiO₂/CA-9.3 can account for its higher HO˙ generation ability than that of TiO₂/CA-3.1 with hydrophobic surface. However, it is still puzzling that the HO˙ amount produced in TiO₂/CA-9.3 is even larger than that in pristine TiO₂, concerning the super-hydrophilic TiO₂ vs. hydrophilic TiO₂/CA-9.3. We propose that a structure-related effect exists in TiO₂/CA-9.3 toward HO˙ generation and the mechanism of HO˙ formation on the outer and inner surfaces of TiO₂/CA-9.3 is illustrated in Fig. 8 (vide infra).

Fig. 8 Schematic illustration of DMP degradation in TiO₂/CA-9.3 under light irradiation. The red ring describes the reactions occurring in the secondary pores. The processes include: (I) on the external surface, HO˙ formation with H₂O reduction and oxidation, (II) the photoelectrons generated on external surface migrates to the internal surface through the carbon layer, (III) on the internal surface, the photoelectrons are trapped by O₂ to produce ˙O²⁻, (IV) HO˙ are generated with the participation of H₂O and ˙O²⁻, (V) DMP is oxidized by active HO˙ on both external and internal surface.

It is understandable that the primary factor inhibiting HO˙ formation via either the oxidation way (eqn (5)) or reduction way (eqn (2)–(4)) is the recombination of photoholes and photoelectrons in the photodegradation process. With respect to the special structure of TiO₂/CA-9.3 with TiO₂ dispersed on both inner and outer surfaces, it is possible that the photoelectrons generated on the outer surface transfer to its inner surface through the conductive CA (process II in Fig. 8). This process potentially increases the separation efficiency of photoelectrons and photoholes. As a result, HO˙ formation via H₂O oxidation with photoholes (eqn (5)) is enhanced on the outer surface TiO₂ (process I in Fig. 8). Meanwhile, the formation of HO˙ via O₂ reduction with participation of H₂O and photoelectrons (eqn (2)–(4)) is improved in the hydrophilic secondary channels of TiO₂/CA-9.3 (process IV in Fig. 8 amplifying red-circle region), because in this secondary channels the three integral components of photoelectrons, O₂ and H₂O needed for HO˙ formation are simultaneously satisfied with the photoelectrons transfer aforementioned, air bubbling in aqueous solution, and surface hydrophilic property, respectively. In addition, the spatial confinement effect^44,45 arising from narrower secondary channel further facilitates the HO˙ formation by stabilizing the intermediates of ˙O²⁻ and H₂O₂. In summary, the interface electron transfer between TiO₂ and CA-9.3 promotes the spacial separation of photoelectrons and photoholes, and hydrophilic secondary-pore surface with confined space results in the increase of HO˙ formation possibility in reduction way (eqn (2)–(4)), and thereby leads to the highest DMP degradation activity of TiO₂/CA-9.3 (process V in Fig. 8).

Comparing with TiO₂/CA-9.3 with uniform distribution of TiO₂ on both outer and inner surface of CA-9.3 support, prepared by introducing CA-9.3 into titanium precursor solution prior to sol–gel formation, TiO₂ + CA-9.3 is a simply mixture of TiO₂ and CA-9.3 and no hydrophilic porous structure exists, in addition, part of CA-9.3 covering on TiO₂ could block the incident light by its self-absorption. Therefore, the degradation activity of TiO₂ + CA-9.3 is far lower than TiO₂/CA-9.3, even lower than TiO₂/CA-3.1 and TiO₂/CA-4.3, where TiO₂ mainly dispersed on the outer surface of CA supports as their small pore diameter.

Conclusions

TiO₂/CA-X photocatalysts with variable pore structures were fabricated and the correlation between the pore structure and the photocatalytic property was systematically studied. Our date indicates that dispersion of TiO₂ on CA-X relies on the pore diameter of CA support. The large diameter enables TiO₂ to be dispersed on both inner and outer surface of CA support, accompanying with the formation of the secondary pore with narrower channel. This special porous structure of TiO₂/CA-9.3 leads to the separation of electrons and holes in space in the photocatalytic process. Correspondingly, the two routes of HO˙ formation in H₂O oxidation and O₂ reduction with participation of H₂O are promoted. The largest amount HO˙ finally causes the highest DMP degradation efficiency in TiO₂/CA-9.3. It was concluded that the surface/interface synergy effect plays a positive role on the spatial separation and utilization of electrons and holes in photocatalytic process, which suggests a potential strategy for designing high efficiency photocatalysts.

Experimental

Materials

Titanium(IV) isopropoxide (95%) and tetraethyl orthosilicate (TEOS, 98%) were purchased from Aladdin (Shanghai, China). Other reagents used were all analytical reagents and purchased from Guangzhou Chemical Reagent Plant (Guangzhou, China). All reagents were used as received without further purification, and the double distilled water was used in all experiments.

Preparation

The carbon aerogels were synthesized following a previously published procedure.³¹ Briefly, the raw materials of glucose (2 g), tetraethyl orthosilicate (TEOS, 3 mL) and sulfuric acid (H₂SO₄) solution (4 mL, pH = 2) were mixed and stirred until a homogenate was formed. The mixture was then treated by hydrofluoric acid (HF, 4 wt%) under stirring, gelated, and aged in a plastic bottle as follows: 40 °C for 2 days, followed by 100 °C for 6 h, and finally 160 °C for 6 h. The xerogel was subsequently carbonized in nitrogen atmosphere at 900 °C for 3 h, and washed by 40 wt% HF to remove silica. The HF/TEOS molar ratio used was 0, 1/30 and 1/7, and as a result, the pore diameters were controlled by changing the amount of HF during the sol–gel process of TEOS. The notation of “CA-X” is used to denote the samples, with X being the pore diameter of CA calculated by the density function theory (DFT) model. The TiO₂/CA-X sample was synthesized with CA-X as the support material.

TiO₂/CA-X samples were prepared by a sol–gel method with titanium isopropoxide as Ti source. A typical procedure is as follows. Titanium(IV) isopropoxide (3 mL) was introduced into ethanol (20 mL) with vigorous stirring for 30 min, and then concentrated nitric acid (0.16 mL) was added into this solution. After stirring for 5 minutes, CA (0.16 g) was distributed into this solution. The mixture solution was constantly stirred until a homogenous CA-contained gel was formed. The gel was aged in air to form a ergoes, which was then ground into fine powder. The powder was calcined at 400 °C for 5 h in the nitrogen atmosphere, and TiO₂/CA-X samples were thus obtained. For comparison, pure TiO₂ was also prepared following the same procedure in the absence of CA. Mechanically mixed TiO₂ and CA-X sample with the same mass ratio as TiO₂/CA-X samples, marked as TiO₂ + CA-9.3 was also prepared.

Characterization

The crystalline structures of TiO₂/CA-X samples were analyzed by D8 Advance X-ray Diffractometer (XRD, Bruker, D8 advance, Germany) with Cu-Kα radiation source at scanning rate of 10 min⁻¹. The morphologies of TiO₂/CA-X samples were observed by transmission electron microscope (TEM, JEOL, JEM-2100, Japan) on a JEM 2100F 200 kV field emission transmission electron microscope. The carbon content in TiO₂/CA-X samples was determined by thermogravimetry analysis (TG, Netzsch, TG209, Germany) in air atmosphere. The pore structures of samples were probed by N₂ adsorption–desorption isotherm at 77 K (Quantachrome, AutosorbiQ, American). Before measurement, the samples were out-gassed at 200 °C for 12 h. The specific surface area was determined by Brunauer–Emmett–Teller (BET) model, while the pore size and volume were calculated by density functional theory (DFT) method on Quantachrome ASiQwin software. Diffuse reflection spectra (DSR) were recorded on UV-vis spectrophotometer (Shimadzu, UV-3150, Japan, equipped with an integrating sphere) to study the optical properties of the samples.

Electron Paramagnetic Resonancse spectrometer (EPR, Bruker, A 300, Germany) were used to detect the hydroxyl radical with 5,5-dimethy-1-pyrroline-N-oxide (DMPO, Sigma-Aldrich) as the radical trap. The samples were irradiated with xenon arc source system (LOT Oriel 75-150, UK). Data acquisition parameters were as follows: magnetic field, 3514 G; microwave power, 5.95 mW; modulation frequency, 100 kHz; modulation amplitude, 1.00 G; microwave frequency, 9.86 GHz; sweep time, 20.48 s. The hydrophilic properties of CA-X and TiO₂/CA-X were measured by OCA contact angle system (Dataphysics, OCA20, Germany). The intermediates in the degradation process were identified from positive modes of MS (AB Sciex, triple QuadTM 4500).

Adsorption and photocatalytic degradation of DMP

Dimethyl phthalate (DMP, 2 mg L⁻¹) was chosen as the model target to investigate the photocatalytic properties of porous TiO₂/CA-X samples. Prior to the photocatalytic reaction, the adsorption experiment of DMP on TiO₂/CA-X was carried out in dark at room temperature by adding 0.01 g of TiO₂/CA-X sample into 100 mL 2 mg L⁻¹ DMP aqueous solution. The photodegradation experiments were carried out in a cylindrical glass reactor (Φ 9.0 cm × 10.6 cm) covered by a cylindrical quartz vessel (Φ 9.0 cm × 9.0 cm), which was full-filled with double distilled water to remove the infrared light radiated from 300 W Xe lamp (Changtuo, PLS-SXE 300/300 UV, Beijing). The light intensity was 85 mW cm⁻² determined by a radiometer (FZ-A, Photoelectric Instrument factory of Beijing Normal University). The emission spectrum of the Xe lamp has been reported in ref. 8. For a typical photocatalytic experiment, 0.1 g L⁻¹ of TiO₂/CA-X catalysts were firstly pre-adsorbed with 100 mL 2 mg L⁻¹ DMP in dark before irradiation to establish adsorption–desorption equilibrium. Then photocatalytic experiments were initiated by air bubbling and irradiation. During the experiments, 0.5 mL of solution was sampled at preset time interval for the analysis of the concentration of DMP by high performance liquid chromatography (HPLC, Techcomp, LC2130, Shanghai, China) equipped with a reverse phase column (Waters, XT erra MS C-18, 5 μm) and a UV detector. The mobile phase was composed of 50% acetonitrile and 50% water, and the detection wavelength was 276 nm. For comparison, pristine TiO₂ and mechanical mixtures of TiO₂ and CA-9.3 (TiO₂ + CA) were also tested under the same condition. The amount of TiO₂ was to the same as that in TiO₂/CA-X.

Acknowledgements

The research was financially supported by the 973 Program (2014CB845600), NSFC (No. 21103235), NSF of Guangdong Province (No. S2012010010775), the Key Laboratory of Fuel Cell Technology of Guangdong Province, and the Key Laboratory of Environmental Pollution Control and Remediation Technology of Guangdong Province (No. 2011K0011). JZZ is grateful to the BES Division of the US DOE for financial support.

References

1. Y. C. Chung and C. Y. Chen, Water, Air, Soil Pollut., 2009, 200, 191–198.
2. Y. H. Chen, L. L. Chen and N. C. Shang, J. Hazard. Mater., 2009, 172, 20–29.
3. W. C. Liao and P. Wang, J. Braz. Chem. Soc., 2009, 20, 866–872.
4. W. J. Jiang, J. A. Joensa, D. D. Dionysiou and K. E. O' Shea, J. Photochem. Photobiol., A, 2013, 262, 7–13.
5. Z. J. Huang, P. X. Wu, Y. H. Lu, X. R. Wang, N. W. Zhu and Z. Dang, J. Hazard. Mater., 2013, 246–247, 70–78.
6. Y. Jing, L. S. Li, Q. Y. Zhang, P. Lu, P. H. Liu and X. H. Lu, J. Hazard. Mater., 2011, 189, 40–47.
7. L. Xu, X. Yang, Y. H. Guo, F. Y. Ma, Y. N. Guo, X. Yuan and M. X. Huo, J. Hazard. Mater., 2010, 178, 1070–1077.
8. J. Y. Shi, H. N. Cui, Z. X. Liang, X. H. Lu, Y. X. Tong, C. Y. Su and H. Liu, Energy Environ. Sci., 2011, 4, 466–470.
9. D. Zhao, C. C. Chen, Y. F. Wang, H. W. Ji, W. H. Ma, L. Zang and J. C. Zhao, J. Phys. Chem. C, 2008, 112, 5993–6001.
10. S. G. Bairu, E. Mghanga, J. Hasan, S. Kola, V. J. Rao, K. Bhanuprakash, L. Giribabu, G. P. Wiederrecht, da. R. Silva, L. G. C. Rego and G. Ramakrishna, J. Phys. Chem. C, 2013, 117, 4824–4835.
11. M. Garcia-Mota, A. Vojvodic, F. Abild-Pedersen and J. K. Norskov, J. Phys. Chem. C, 2013, 117, 460–465.
12. R. Leary and A. Westwood, Carbon, 2011, 49, 741–772.
13. C. Pelekani and V. L. Snoeyink, Water Res., 1999, 33, 1209–1219.
14. X. W. Zhang, M. H. Zhou and L. Lei, Carbon, 2005, 43, 1700–1708.
15. Y. H. Ao, J. J. Xu, D. G. Fu, X. W. Shen and C. W. Yuan, Colloids Surf., A, 2008, 312, 125–130.
16. H. Zhang, X. J. Lv, Y. M. Li, Y. Wang and J. H. Li, ACS Nano, 2010, 4, 380–386.
17. H. Y. Ng, S. Ikeda, M. Matsumura and R. Amal, Energy Environ. Sci., 2012, 5, 9307–9318.
18. W. D. Wang, C. G. Silva and J. L. Faria, Appl. Catal., B, 2007, 70, 470–478.
19. W. Wei, C. Yu, Q. F. Zhao, G. S. Li and Y. Wan, Chem.–Eur. J., 2013, 19, 566–577.
20. Y. Yao, G. Li, S. Ciston, R. M. Lueptow and K. A. Gray, Environ. Sci. Technol., 2008, 42, 4952–4957.
21. X. H. Xia, Z. H. Jia, Y. Yu, Y. Liang, Z. Wang and L. L. Ma, Carbon, 2007, 45, 717–721.
22. L. Gu, Z. X. Chen, C. Sun, B. Wei and X. Yu, Desalination, 2010, 263, 107–112.
23. M. F. Wu, Y. N. Jin, G. H. Zhao, M. F. Li and D. M. Li, Environ. Sci. Technol., 2010, 44, 1780–1785.
24. D. Wang, G. H. Yang, Q. X. Ma, M. B. Wu, Y. S. Tan, Y. Yoneyama and N. Tsubaki, ACS Catal., 2012, 2, 1958–1966.
25. H. Q. Yang, L. Zhang, L. Zhong, Q. H. Yang and C. Li, Angew. Chem., Int. Ed., 2007, 46, 6861–6865.
26. Z. J. Chen, X. H. Li and C. Li, Chin. J. Catal., 2011, 32, 155–161.
27. W. Chen, Z. L. Fan, B. Zhang, G. J. Ma, K. Takanabe, X. X. Zhang and Z. P. Lai, J. Am. Chem. Soc., 2011, 133, 14896–14899.
28. Y. N. Jin, M. F. Wu, G. H. Zhao and M. F. Li, Chem. Eng. J., 2011, 168, 1248–1255.
29. J. M. Miller and B. Dunn, Langmuir, 1999, 15, 799–806.
30. H. D. Du, L. Gan, B. H. Li, P. Wu, Y. L. Qiu, F. Y. Kang, R. W. Fu and Y. Q. Zeng, J. Phys. Chem. C, 2007, 111, 2040–2043.
31. D. C. Wu, Z. H. Li, Y. R. Liang, X. Q. Yang, X. H. Zeng and R. W. Fu, Carbon, 2009, 47, 916–918.
32. U. Heudorf, V. Mersch-Sundermann and E. Angerer, Int. J. Hyg. Environ. Health, 2007, 210, 623–634.
33. S. Takeuchi, M. Iida, S. Kobayashi, K. Jin, T. Matsuda and H. Kojima, Toxicology, 2005, 210, 223–233.
34. Z. Y. Ryu, J. T. Zheng, M. Z. Wang and B. J. Zhang, Carbon, 1999, 37, 1257–1264.
35. M. L. Occelli, J. P. Olivier, J. A. Perdigon-Melon and A. Auroux, Langmuir, 2002, 18, 9816–9823.
36. J. H. Park, S. K. Kim, H. S. Kim, Y. J. Cho, J. Park, K. E. Lee, C. W. Yoon, S. W. Nam and S. O. Kang, Chem. Commun., 2013, 49, 10832–10834.
37. C. Donati, M. Drikas, R. Hayes and G. Newcombe, Water Res., 1994, 28, 1735–1742.
38. P. Pendleton, R. Schumann and S. H. Wong, J. Colloid Interface Sci., 2001, 240, 1–8.
39. A. L. Linsebiger, G. Q. Lu and J. T. Yates Jr, Chem. Rev., 1995, 95, 735–758.
40. S. Sakthivel, M. C. Hidalgo, D. W. Bahnermann, S. U. Geissen, V. Murugesan and A. Vogelpohl, Appl. Catal., B, 2006, 63, 31–40.
41. J. H. Im, S. J. Yang, C. H. Yun and C. R. Park, Nanotechnology, 2012, 23, 035604.
42. M. R. Hoffmann, S. T. Martin, W. Choi and D. F. Bahnemann, Chem. Rev., 1995, 95, 69–96.
43. J. H. Sun, X. L. Wang, J. Y. Sun, R. X. Sun, S. P. Sun and L. P. Qiao, J. Mol. Catal. A: Chem., 2006, 260, 241–246.
44. J. Shi, Chem. Rev., 2013, 113, 2139–2181.
45. I. J. Morales, J. S. González, P. M. Torres and A. J. López, Appl. Catal., B, 2011, 105, 199–205.

---

Footnote

† Electronic supplementary information (ESI) available: The thermogravimetry (TG), differential thermal gravity (DTG), X-ray diffraction and UV-vis absorption spectra for all samples. See DOI: 10.1039/c6ra08074j

---