Ind. Eng. Chem. Res. 2009, 48, 2891–2898
2891
Buoyant Photocatalyst with Greatly Enhanced Visible-Light Activity Prepared through a Low Temperature Hydrothermal Method Hui Han and Renbi Bai* DiVision of EnVironmental Science & Engineering, Faculty of Engineering National, UniVersity of Singapore, Singapore 117576
In this study, TiO2 was modified and immobilized on polypropylene (PP) granules to prepare buoyant photocatalyst with visible light activity. The TiO2 nano-sol was first prepared from titanium n-butoxide (Ti(OBu)4) in the presence of acetyl acetone (AcAc) or acetic acid (AcOH) as an inhibiting agent and subsequently treated with triethylamine (TEA) under the condition of room temperature. Then, a simple onestep process was utilized for the simultaneous crystallization and immobilization of the treated TiO2 nanoparticles on the PP substrate in a low temperature hydrothermal reactor (150 °C). It was found that the TEA-treated TiO2 nano-sol with AcAc as the inhibiting agent led to high visible light activity for the prepared photocatalyst from the low temperature hydrothermal process. The light absorption edge was up to 800 nm, and the visible light absorption rate reached 32 to 66%, depending on the TEA-treatment time. XRD analysis showed all the TiO2-based photocatalysts having the major crystal structure of anatase and minor of brookite. Raman spectroscopy, however, clearly revealed the existence of the nitrogen-doped composition (i.e., TiO2-xNx) in the photocatalyst with high visible light activity. Experiments conducted for decolorization of methyl orange (MO) dye solutions showed the effectiveness of the prepared novel photocatalyst under both the UV and the visible light. 1. Introduction TiO2 has been well established as the choice of photocatalysts studied for the degradation of aqueous organic pollutants under the UV light. This is, to a large extent, attributed to the advantages of TiO2 having relatively high photoactivity under the UV light, being nontoxic and readily available in the market. Traditionally, TiO2 photocatalyst is produced in the form of powder and is applied directly in a suspended form into the solution to be treated during the various photocatalytic applications. Recent research has also established the fact that TiO2 photocatalyst in the nanoparticle form is much more active than that in the larger sized powder. However, the application potential of the nanosized photocatalysts in water or wastewater treatment has been greatly limited by the fact that the separation of the photocatalyst nanoparticles from the water being treated is usually difficult and can incur high extra cost.1 To overcome the problem for separation, there have been a number of research efforts made to immobilize TiO2 nanoparticles on various larger substrates, particularly the inorganic ones, including glass, metals, silicon slides, and even stones.2,3 On the other hand, as TiO2 photocatalyst is only active under the UV light (200-400 nm), artificial UV light sources usually have to be provided for almost all the photocatalytic applications. Unfortunately, light, including UV light, attenuates significantly in water with the traveling distance, especially in wastewater that may contain a high content of particles (the attenuation coefficient in water is at least 100 times greater than that of air4). As a result, the light supplied to the photocatalytic reactions in water often has a low light utilization efficiency.5,6 To improve the efficiency, especially for the expensive UV light provided, quartz reactors with very short light traveling distance in water have been developed for the TiO2 photocatalyst. These reactors usually consist of many thin gap water flow channels, with UV light lamps installed along the channel walls. Although the light * To whom correspondence should be addressed. E-mail: esebairb@ nus.edu.sg. Phone: +65 6516 4532. Fax: +65 6774 4202.
utilization efficiency may be improved, the process incurs very high capital and operational costs.1,6 To make the photocatalytic technology cost-competitive for practical applications in water or wastewater treatment, significant recent research interest has been directed to utilize the naturally available sunlight as a possible “free” light source for photocatalytic reactions. A common approach to achieve this has been to modify TiO2 to extend its light activity to the visiblelight range (400-700 nm)7-9 because the natural sunlight only contains the UV light at as low as 3-5% but visible light up to 45-50%.10 Most of the modification studies have been carried out by doping TiO2 with precious metals, metal oxides, or inorganic components such as nitrogen (N) or carbon (C) and so forth that can reduce the band gap energy of the photocatalyst. Commonly, almost all of these studies involved at least a high temperature treatment process in the preparation method (often >300-400 °C).11-14 In the literature, there have been reports using hollow glass microspheres as a substrate for the immobilization of TiO2 photocatalyst particles. These studies aimed at either improving the separation performance of the photocatalyst particles after their use or making the photocatalyst floatable in water for enhanced UV light utilization.15-17 The preparation method typically involved a dip-coating of the hollow glass microspheres in a TiO2 sol, followed by a calcination treatment of the coated microspheres at a high temperature (e.g., above 400 °C).18 A major limitation in these developments lies in the fact that the hollow glass microspheres are generally expensive and, as a substrate, they are fragile and can easily break, especially in the high temperature calcination process. In this paper, we report a new development to prepare buoyant photocatalyst with high visible light activity on a polypropylene (PP) substrate. The objective is to obtain a photocatalyst that can be used with the sunlight, directly at the water surface, for photocatalytic oxidation of organic contaminants in water or wastewater. The novel photocatalyst was obtained through modifying TiO2 by a doping treatment and then crystallizing
10.1021/ie801362a CCC: $40.75 2009 American Chemical Society Published on Web 02/13/2009
2892 Ind. Eng. Chem. Res., Vol. 48, No. 6, 2009
and immobilizing the treated TiO2 on PP granules. PP was selected as the substrate because it has a much lower density of 850-950 kg/m3 than water and hence is highly floatable. PP has also been known to have high resistance to lights including UV19 and good chemical stability in acid/base and many organic solvents.20 In contrast with hollow glass microspheres as an immobilization substrate, PP is also cheap and highly durable with good mechanical strength. However, the use of PP as the immobilization substrate has restricted the preparation process to be necessarily conducted at a relatively low temperature because PP cannot physically stand a reaction temperature higher than 160 °C. This condition has made the commonly established methods that usually need a high temperature for crystallization, doping modification, or immobilization of TiO2 photocatalyst (e.g., above 300-400 °C)11-14 inapplicable. To confine the preparation method to a low temperature process, the three possible methods for selection may include plasma enhanced chemical vapor deposition (PE-CVD), liquid phase deposition (LPD), and hydrothermal reaction. The PE-CVD method in general requires very expensive devices.21 It has been reported that TiO2 photocatalyst prepared by the LPD method has very low crystallinity and hence low photoactivity, unless calcinated at 300 °C,11-14 especially to prepare doped TiO2 for visiblelight activity. Although TiO2 photocatalyst was prepared through a hydrothermal method at a temperature as low as 150 °C,22 almost all the TiO2 photocatalysts with visible light activity required a doping reaction at temperatures at least 180 °C or above.8,9,23-25 Nevertheless, Gole et al.26 have reported a study to prepare nitrogen-doped crystal TiO2 photocatalyst at room temperature through treating TiO2 nano-sol with triethylamine (TEA) in the presence of a large amount of palladium salt as the phase transfer catalyst. The successful doping of nitrogen in TiO2 at room temperature was ascribed to the small size of the TiO2 nano-sol particles. However, as confirmed by our preliminary study, the TiO2 nano-sol cannot be turned into crystallined TiO2 photocatalyst after TEA treatment without using the expensive precious metal salt as the phase transfer catalyst. Their preparation method has however inspired us to dope nitrogen into TiO2 lattice through TiO2 nano-sol at a relatively low temperature without the use of precious metal catalyst to obtain visible light activity. To accomplish this idea, we treated TiO2 nano-sol with TEA in the presence of an organic inhibiting agent and then simultaneously crystallized and immobilized the modified TiO2 on the PP substrate in a low temperature hydrothermal reactor (150 °C). 2. Experimental Section 2.1. Material Preparation. The TiO2 nano-sol was first prepared in three steps: (a) 5 mL of titanium n-butoxide (Ti(OBu)4, 99%, Sinopharm) was dissolved into 23 mL of absolute ethanol (99%, Merck), followed by the addition of 2 mL of glacial acetic acid (AcOH, 99.8%, Merck) or 0.5 mL of acetyl acetone (AcAc, 99%, Sinopharm) into the Ti(OBu)4 solution as an inhibiting agent for the control of the hydrolysis rate of Ti(OBu)4; (b) a solution made of 1.5 mL of ultra-pure water, 1.5 mL of nitric acid solution (0.1 mol/L), and 12 mL of absolute ethanol was then added dropwise into the Ti(OBu)4 solution at a rate of 0.5-1 mL/min; and (c) the mixture was aged overnight with stirring for 12 h. The process produced TiO2 nano-sol containing particles at a size around 5-6 nm as revealed from the TEM analysis. Subsequently, the TiO2 nanosol was treated with triethylamine (TEA, 99%, Sinopharm) under room temperature (23-24 °C) by adding 4.1 mL of TEA into the sol with stirring. The treatment time was 12 h in most
cases initially but also varied to a much longer time (48, 120, or 160 h) in some later experiments. After the treatment with TEA, the TiO2 nano-sol was heated at around 85 °C in a fume hood until it turned into dry gel or powder. (In this paper, TEAtreated photocatalysts with AcOH or AcAc as the inhibiting agent will be labeled separately because of their different properties. Photocatalysts prepared with either AcOH or AcAc as the inhibiting agent but not treated with TEA will be denoted as “TiO2 without TEA treatment” and not be further differentiated because they did not show any differences in the TEM or FESEM observation, XRD patterns, Raman spectra, XPS, UV-vis spectra, and photocatalytic activity.) The crystallization and immobilization of the TEA-treated TiO2 on PP were achieved simultaneously in a 900 mL hydrothermal reactor (Berghof Br900, Germany). Around 840 pieces of PP granules with a size of around 2 × φ3.5 mm (length × diameter) and a total weight of 18.2 mg (pretreated in 100 mL of K2Cr2O4/H2SO4 solution for 5 min at room temperature) were added, together with the TEA-treated TiO2 dry gel and 150 mL of ultra-pure water, into the hydrothermal reactor. The reactor was placed on a hot plate magnetic stirrer (Heidolph) equipped with an aluminum heating block. The contents in the reactor were stirred with a PTFE magnetic stirrer bar at 250 rpm and were heated for 30 min to reach 150 °C. The temperature was then maintained at 150 °C for 10 h to allow the reaction process to proceed sufficiently. After that, the reactor was cooled to room temperature naturally. The PP granules immobilized with a film of the photocatalyst nanoparticles were collected from the reactor, washed sufficiently with DI water, and then dried in an oven (65 °C) to a constant weight for further use. The photocatalyst particles in the solution were also collected through centrifugation, washed with ethanol and DI water, and then dried similarly as mentioned above for further use. 2.2. Material Characterizations. The particle dimension of the photocatalyst powder produced in the reactor was observed with a transmission electron microscope (JEOL JEM-2010, Japan) operated at 200 kV. The surface morphology of the photocatalyst film on the PP granules was observed with a field emission scanning electron microscope (JEOL JSM-6700F, Japan) under a 5 kV electron beam. The specific surface area of the photocatalyst powder was measured through BET analysis (Quantachrome Nova 4200e, U.S.A.). The crystal structures of the prepared photocatalyst samples were examined with a powder X-ray diffractometer (LabX XRD-6000, Japan), and the crystal compositions of the samples were analyzed through Raman spectroscopy (Bruker 106/S attached to Bruker EQUINOX55 FT-IR spectrometer, U.S.A.) with a 1064 nm Nd:YAG laser source. XPS analysis was conducted to verify the states and species of elements of interest in the modified TiO2 photocatalyst film on the PP granules through an X-ray photoelectron spectrometer (Kratos XPS system-AXIS His-165 Ultra, U.K.) with an Al K radiation source (1486.7 eV). UV-vis reflectance spectra for the samples in powder form were obtained through a φ60 mm integrating sphere accessory attached to the UV-vis spectrophotometer (Jasco V660, Japan) to reveal their light activity (absorbance). 2.3. Photocatalytic Activity Tests. The photocatalytic activity of the prepared photocatalysts was evaluated through the decolorization experiments for methyl orange (MO) dye in aqueous solutions. A 100 mL PYREX beaker with a diameter of 50 mm was filled with 50 mL of MO dye solution at an initial concentration of 15 mg/L. The light from a 150 W xenon lamp (Newport, U.S.A.) with a 69 mm diameter radiation area
Ind. Eng. Chem. Res., Vol. 48, No. 6, 2009 2893
was applied to the MO solution in the beaker. The UV light power was around 48 W/m2, and the visible light was 178 W/m2 from the light source. For photocatalytic reaction under the visible light alone, a long-wave pass filter with a “cut-on wavelength” at 400 nm was used to block the UV light. The filter was mounted on the top of the beaker, and the content in the beaker was stirred with a magnetic stirrer bar at 180 rpm. The distance from the liquid surface in the beaker to the light source was 160 mm. Initially, the light activity of the prepared photocatalysts was tested with the powder samples collected from the solution for convenience and easy characterization in the photocatalyst preparation. A 40 mg amount of the photocatalyst powder samples was added into the MO dye solution. The photocatalytic oxidation reactions under the full irradiation of the 150 W xenon lamp (without UV blocking) were labeled as “UV-vis” in the following discussion and were conducted for 20 min. Those reactions under the visible light alone (with UV blocked) were referred to as “Vis” hereafter and were conducted for 100 min. After a decoloration test, the contents in the beaker were centrifuged to separate the photocatalyst particles, and the supernatant was taken to analyze the remaining MO dye concentration with a UV-vis spectrophotometer at 465 nm (Jasco V660, Japan). The absorbance readings were converted to the weight concentrations through a prepared calibration curve. Finally, the buoyant photocatalyst with TEA-treated TiO2 particles crystallized/immobilized on the PP granules were tested in the decolorization experiments. The photocatalyst was prepared with a TEA treatment time of 12 h and AcAc as the inhibiting agent in this case. A total of 100 PP granules with about 15-20 mg of the photocatalyst particles immobilized on the surfaces (prepared through three immobilization cycles) were added into the PYREX beaker. The buoyant photocatalyst granules formed a floating layer on the surface of the MO dye solution. Other experimental conditions were the same as described earlier for the powder photocatalyst. Because a less amount of photocatalyst was used, the decolorization experiments were conducted for up to 4 h. Samples of about 2 mL were collected at each hour from the beaker by a plastic dropper and analyzed with the UV-vis spectrophotometer for the MO dye concentration. After each analysis, the sample solution was returned to the reaction beaker to maintain the same volume of solution in the reaction. 3. Results and Discussion 3.1. Morphologies of Prepared Photocatalysts. Figure 1 shows the microscopic images of the prepared TiO2 photocatalyst particles collected from the solution and the surfaces of the PP granules before and after the immobilization of the TEAtreated TiO2 nanoparticle film. The photocatalyst particles produced in the solution have a more or less ellipse shape with a dimension at around 7 × 10 nm, as shown in Figure 1a. BET analysis indicated a specific surface area of about 169 m2/g for those particles. The photocatalyst film on the PP granule appears to consist of more spherically shaped particles with a size up to 30 nm and possibly is in multilayers; see Figure 1d. Thus, the prepared photocatalyst particles grown on the PP substrate have a significantly larger size than those grown in the solution. This may be attributed to the fact that the powder particles in the solution were produced primarily through a homogeneous nucleation mechanism while those on the PP granules were produced through a heterogeneous nucleation mechanism that requires less Gibbs free energy and can thus start nucleation earlier than the homogeneous nucleation.27 As a result, the
Figure 1. Surface morphologies: (a) TEM image of TEA-treated TiO2 (in the presence of AcAc) as powder particles; (b) FESEM image of blank PP granule; (c) FESEM image of TEA-treated TiO2 (in the presence of AcAc) immobilized on PP granules (×15 000); (d) FESEM image of TEA-treated TiO2 (in the presence of AcAc) immobilized on PP granules (×50 000).
Figure 2. XRD patterns: (a) TiO2 without TEA treatment; (b) TEA-treated TiO2 inhibited by AcOH; and (c) TEA-treated TiO2 inhibited by AcAc.
crystallization time of the photocatalyst particles on PP was longer than those in the solution, leading to a larger particle size in the film. However, microscopic observations on various samples did not seem to indicate that the morphologies of the photocatalyst particles were noticeably affected by the TEA treatment of the TiO2 nano-sol. 3.2. Crystalline Structures and Compositions of Prepared Photocatalysts. XRD analysis was conducted for the photocatalyst samples prepared without TEA treatment and with TEA treatment in the presence of AcOH or AcAc as the inhibiting agent. As shown in Figure 2, all three samples exhibit very similar XRD patterns with the main component of anatase (A peaks) and minor of brookite (B peaks). The XRD results hence do not seem to show noticeable differences in the crystal structure of the prepared TiO2 photocatalysts because of the TEA treatment on the TiO2 sol. The Raman spectra of the three samples were further obtained, as shown in Figure 3. In Figure 3a, the peaks at 147, 403, 516, and 641 cm-1 in all three samples can be attributed to the characteristic peaks of the anatase phase, indicating anatase to be the major component, which agreed with the XRD results in Figure 2. Although little difference could be observed in the Roman spectra for the sample without TEA treatment and that with TEA treatment at the presence of AcOH as the inhibiting agent, the sample with TEA treatment at the presence of AcAc clearly showed enhanced peaks at 403,
2894 Ind. Eng. Chem. Res., Vol. 48, No. 6, 2009
Figure 3. Raman Spectra: (a) overall for untreated and TEA treated TiO2; (b) fitted curves for TEA treated TiO2 with AcAc as the inhibiting agent.
Figure 4. XPS spectra of (a) Ti 2p peak; (b) O 1s peak; and (c) N 1s peak for TEA-treated TiO2 film on PP granule with AcAc as the inhibiting agent.
516, and 641 cm-1. Therefore, the Raman spectrum of this sample was further analyzed with the peak fitting module from Origin (with FFT filter on the basis of Gaussian peak fit calculus by the software of OriginPro 7.5). As indicated by the results shown in Figure 3b, the model fitting analysis clearly reveals the presence of the titanium nitride (TiO2-xNx) structure in the sample. Although the peaks at 403, 516, and 641 cm-1 are usually assigned to the anatase features, those at 330, 580, and 690 cm-1 can be attributed to the titanium nitride structure.28 The existence of the TiO2-xNx structure in the sample enhanced the peaks at 403, 516, and 641 cm-1 for the anatase structure. Thus, the Raman spectroscopic analysis provides evidence that the TEA treatment for the TiO2 nano-sol with AcAc as the inhibiting agent has resulted in the prepared photocatalyst being doped with nitrogen atoms. 3.3. XPS Studies. XPS analysis was also conducted for the various samples. The XPS survey scans (see Supporting Information, Figure S1) always showed that blank PP granules mainly had the C 1s peak (very strong), and the PP granules immobilized with the TEA-treated TiO2 photocatalyst had Ti 2p (strong) and Ti 2s (weak) peaks as well as the O 1s peak (very strong). These results confirm the effective immobilization of the prepared photocatalyst on the PP granules in the low
temperature hydrothermal process. The XPS spectra of Ti 2p, O 1s, and N 1s for the photocatalyst film on the PP granules were further examined. As shown in Figure 4 a, the Ti 2p peaks are similar to those of the standard TiO2 sample reported,29 and the O 1s main peak at 529.5 eV in Figure 4b can be assigned to the oxygen in the TiO2 lattice. The N 1s spectrum in Figure 4c shows a small but distinguishable peak at 399.4 eV. This peak can be assigned to the nitrogen in the O-Ti-N linkages28,30 and could not arise from any nitrogen from the adsorbed TEA on the surface that would give a N 1s peak at 395.4 eV.31 Hence, the XPS results in Figure 4 confirm that the immobilized photocatalyst film on the PP granules was composed of TiO2 and had nitrogen atoms being doped into the TiO2 lattices, attributed to the TEA treatment with AcAc as the inhibiting agent. The result from the XPS analysis agreed with the results revealed by the Raman spectroscopic study mentioned early. The small peak of N 1s in Figure 4c is expected because the amount of nitrogen doped was essentially at a trace level. However, the small peak of N 1s was not observed for the photocatalyst film on the PP granules prepared with AcOH as the inhibiting agent. It was also found from the analysis that the XPS spectra for the photocatalyst films on the PP granules were the same as those for the photocatalyst particles collected
Ind. Eng. Chem. Res., Vol. 48, No. 6, 2009 2895
Figure 5. UV-vis absorption spectra for TiO2 powder samples without TEA treatment (denoted as TiO2), with TEA treatment at the presence of AcOH (denoted as AcOH) or AcAc (denoted as AcAc).
literature for nitrogen-doped TiO2 photocatalysts to our best knowledge.32 Thus, the inhibiting agent used before the TEA treatment had an effect on the nitrogen-doping extent and hence the light activity of the prepared photocatalysts. The nitrogendoping effect (which contributes to the lower band gap energy of the prepared photocatalyst and hence higher visible light absorption rate) that occurred for TEA-treated TiO2 nano-sol with AcAc as the inhibiting agent may be attributed to the different inhibiting mechanism or strength of AcAc and AcOH for Ti(OBu)4. Ti(OBu)4 is known to be easily hydrolyzed in the presence of water and rapidly condensed to form a networked structure,33 as shown in eqs 1 and 2. This networked structure makes it difficult for nitrogen from TEA to be doped into the TiO2 lattice (i.e., a nitrogen atom replaces an oxygen atom).
Table 1. Absorption Ratio to UV Light and Visible Light and Band Gap Energy for TiO2 Powder Samples without TEA Treatment or with TEA Treatment in the Presence of AcOH or AcAc Being Used as the Inhibiting Agenta light absorption light absorption band in UV in vis gap, (200-400 nm), % (400-700 nm), % eV TiO2 without TEA treatment TiO2 with TEA treatment and AcOH as inhibiting agent TiO2 with TEA treatment and AcAc as inhibiting agent a
89.75 88.63
10.54 6.11
2.85 2.94
93.35
31.98
2.49
TEA treatment time ) 12 h.
in the solution from the reactor under each condition. This is probably attributed to the multiple layer composition of the photocatalyst film formed on the PP granules. Even though the initial nucleation mechanism for the TiO2 nanoparticles on the PP surfaces may be different from that for the TiO2 nanoparticles in the solution, the nucleation mechanism for the next layer of TiO2 nanoparticles in the film on the PP surfaces may become the same as that for the TiO2 nanoparticles in the solution. As a result, similar XPS spectra were observed for the photocatalyst film on the PP granules and the photocatalyst particles in the solution from the same experiments because it is well-known that XPS is a surface analysis tool with penetration depth usually less than a few nanometers. One of the implications of those XPS results could be that the prepared photocatalysts in the form of film on the PP granules can have similar light activity as the photocatalytic particles collected in the solution. 3.4. UV-Vis Absorption Spectra. The light absorption rate is an indication of the light activity of a photocatalyst. The UV-vis absorption spectra for the three samples of “TiO2” (denoting TiO2 photocatalyst prepared with the use of either inhibiting agent but without being treated by TEA), “AcOH” (denoting TiO2 photocatalyst prepared with AcOH as the inhibiting agent and being treated by TEA), and “AcAc” (denoting TiO2 photocatalyst with AcAc as the inhibiting agent and being treated by TEA) are shown in Figure 5, and the corresponding light absorption rates are given in Table 1. It is apparent that the modified TiO2 particles through TEA treatment with AcAc being used as the inhibiting agent had a reduced band gap energy and absorbed significantly more light in the visible light range (400-700 nm) than the other two samples. About 32% of the visible light was adsorbed, and the absorption edge was, in fact, even up to 800 nm against the light wavelength. Such a wide absorption range in terms of light wavelength has not often been reported, if not none, in the
The use of AcOH or AcAc as the inhibiting agent was to control the hydrolyzation rate of Ti(OBu)4 and also possibly to interrupt the continuous network structure formation. When one or more of the alkoxide groups in Ti(OBu)4 was replaced through a reaction with coordinating agents, the hydrolyzation rate would be largely reduced. In the case of AcOH being used as the inhibiting agent, the alkoxide groups can be displaced by the AcOH ligands, and a more stable intermediate product was formed; see eq 3.34
This can slow down the hydrolysis and lower the condensation rate of Ti(OBu)4. When AcAc was used as the inhibiting agent, the AcAc molecule can replace an alkoxide group in Ti(OBu)4 and chelate with the titanium atom to form a relatively stable ring structure; see eq 4.
It has been reported that AcAc is much a stronger inhibiting agent than AcOH to reduce the hydrolysis rate of Ti(OBu)4.33 In the TiO2 sol preparation process, we also observed that TiO2 sol inhibited by AcOH would turn into gel in less than a day, in contrast to the case of AcAc being used as the inhibiting agent. A yellow color for the TiO2 nano-sol inhibited by AcAc was observed in the experiment, and it was an indication for the existence of the titanium acetyl acetonate composition, as shown in eq 4. However, the inhibiting agent only slowed down the rate of Ti(OBu)4 hydrolysis but did not eliminate it. Once
2896 Ind. Eng. Chem. Res., Vol. 48, No. 6, 2009
Figure 6. Photocatalytic activity for decolorization of MO dye solutions under the condition of “UV-vis” for 20 min and “Vis” for 100 min respectively by TiO2 powder samples without TEA treatment (noted as TiO2) and with TEA treatment at the presence of AcOH (noted as AcOH) or AcAc (noted as AcAc).
Figure 7. Decolorization of MO dye by buoyant photocatalyst prepared from TEA treated TiO2 (with AcAc being used as the inhibiting agent) and immobilized on PP granules under the condition of “UV-vis” and “Vis” respectively (C0 ) 15 mg/L).
a Ti-OH group is formed, the bond can be attacked by the more reactive nitrogen in TEA, as shown in eq 5.
In the present study, TEA treated TiO2 with AcOH as the inhibiting agent did not result in doping of nitrogen into the TiO2 photocatalyst. This was probably due to the relatively weaker inhibiting ability of AcOH in comparison with AcAc. 3.5. Photocatalytic Oxidation Activity. The activity of the prepared photocatalysts was evaluated through the decolorization experiments for the MO dye solutions. Initially, the powder samples collected from the reactor solution were used. The results obtained under the “UV-vis” condition for 20 min or “Vis” condition for 100 min are presented in Figure 6. Under the “UV-vis” condition, the MO dye solution without the photocatalyst particles (denoted as blank) showed negligible reduction in the concentration. The TEA treated TiO2 photocatalysts prepared with either AcOH or AcAc as the inhibiting agent showed high oxidation activity for the MO dye, with up to 70-80% reduction in the MO dye concentration, as compared to 60% reduction by the photocatalysts without TEA treatment. The results indicated that the TEA-treated TiO2 photocatalyst probably had lower band gap energy and hence showed greater photocatalytic activity. This speculation can be further supported by the results obtained under the “Vis” condition. While the TiO2 photocatalyst without TEA treatment did not show any reduction in the MO dye concentration, the sample prepared through TEA treatment with AcAc as the inhibiting agent significantly reduced the MO dye concentration under the “Vis” condition. In contrast, the photocatalyst prepared through the TEA treatment at the presence of AcOH as the inhibiting agent showed only a very small effect in the reduction of the MO dye concentration. The decolorization of the MO dye solution was further studied with the buoyant photocatalyst (TiO2 was treated with TEA at the presence of AcAc as the inhibiting agent and immobilized on PP granules). The results from a preliminary test are shown in Figure 7. Similar to the results in Figure 6, the reduction of the MO dye concentration was fast under the full light irradiation, that is, under the “UV-vis” condition, and still was very effective under the visible light alone (with UV blocked, labeled as “Vis”). The lower concentration reduction rates in Figure 7 by the buoyant photocatalyst, in comparison with those
Figure 8. Photographs showing the prepared buoyant photocatalyst floating on the solution surface and the color change of the MO dye solution due to the photocatalytic oxidation of the MO dye (a) before the photocatalytic oxidation, (b) after 6 h photocatalytic oxidation, and (c) hourly changes under the “UV-vis” condition. The photocatalyst was TEA-treated TiO2 with AcAc as the inhibiting agent and immobilized on PP granules.
in Figure 6 by the photocatalyst powder particles, are attributed to the significantly lower amount of photocatalyst (about 15 mg immobilized on the PP granules) used in the experiment of Figure 7 and the smaller surface area of the film than that of the powder particles. The photographs in Figure 8 illustrate the prepared buoyant photocatalyst floating on the surface of the MO dye solution and the color changes with the treatment time during the photocatalytic oxidation of the MO dye solution under the “UV-vis” condition from some typical experiments. These preliminary results have demonstrated the feasibility and prospect of using the prepared buoyant photocatalyst for decontamination of organic pollutants in water with the possibility to use sunlight as the effective light source. 3.6. Effect of TEA Treatment Time on Light Activity. It has been found that the treatment time of TEA on the TiO2 nano-sol in the preparation process greatly affected the UV-vis absorption performance and therefore possibly the photoactivity of the prepared photocatalysts. Thus, the modification of TiO2
Ind. Eng. Chem. Res., Vol. 48, No. 6, 2009 2897
Figure 9. UV-vis absorption ratio and MO dye decolorization rate for the prepared photocatalyst with TEA treatment time for 12, 48, and 160 h, respectively, at the presence of AcAc as the inhibiting agent. (a) Light absorption ratio and (b) decolorization under “UV-vis” for 20 min and decolorization under “Vis” for 100 min (C0 ) 15 mg/L). Table 2. Absorption Rate to UV Light and Visible Light and Band Gap Energy for TiO2 Samples Treated with TEA for Different Timesa
12 h 48 h 160 h a
light absorption in UV (200-400 nm), %
light absorption in vis (400-700 nm), %
band gap, eV
93.35 93.84 96.33
31.98 42.99 66.27
2.49 1.89 0.7
With AcAc being used as the inhibiting agent.
photocatalyst through TEA treatment with AcAc as the inhibiting agent was further examined for 12, 48, and 160 h of treatment time, respectively. The light absorption performances of these photocatalyst samples are presented in Figure 9a and Table 2. It is clear that extending the treatment time greatly improved the absorption of the photocatalysts to the visible light. For example, with 160 h of treatment, the photocatalyst achieved an absorption rate for the visible light at as high as 66.27%. A similar phenomenon was also observed by Xu et al.7 who prepared carbon modified TiO2 through hydrolysis of titanium tetrachloride (TiCl4) followed by a treatment with tetrabutylammonium and then a drying and calcination process at 400 and 500 °C, respectively. They attributed the enhanced visible light absorbance to the formation of coke-like carbonaceous species. In our present study, the carbon in TEA may also participate in the doping reaction even although it was not possible to produce the coke-like carbon due to the lower temperature we used. It is interesting to note that extending the TEA treatment time in the preparation process appeared to depress the decolorization rate of the MO dye solutions under “UV” although the decolorization rate was slightly improved under “Vis” (see Figure 9b). In terms of the overall activity under both the UV and visible lights, there may be a compromise to be made to achieve a balanced performance for the prepared photocatalysts through the change of the TEA-treatment time. For full sunlight-driven photocatalytic reaction, however, a longer treatment time by TEA may be advantageous as the solar light contains mostly the visible light. The above observation may also explain the results in Figure 6 where the activity of the AcOH-denoted photocatalyst under “UV-vis” was found to be slightly higher than that of the AcAc-denoted one. Noting that the laboratory light source used in this study provided UV light at around 48 W/m2 and the visible light at 178 W/m2, that is, the UV light was about onefourth of the visible light, in the natural sunlight, however, the UV light is only less than one-tenth of the visible light. Hence, the benefit to enhance the visible light activity of a photocatalyst
for practical applications with sunlight is generally justified, even though the doping reaction may slightly lower the UV light activity. 4. Conclusions Buoyant photocatalyst with visible-light activity was successfully prepared through TEA-treatment of TiO2 nano-sol and subsequent crystallization/immobilization of the nano-sol particles on a PP substrate in a low temperature hydrothermal process (150 °C). The inhibiting agent used in the method was found to have an important effect on the result of TEA treatment and hence the visible light activity of the prepared photocatalysts. The photocatalyst prepared with AcAc as the inhibiting agent achieved greatly enhanced visible-light activity. The light absorption edge was up to 800 nm, and the absorption rate for the visible light reached as high as 66.26%, depending on and increasing with the TEA treatment time. XRD results showed the main crystalline structure of anatase for the photocatalyst. Raman and XPS analyses revealed the doping of the nitrogen atoms into the TiO2 lattices and formed the titanium nitride structure, which lowered the band gap energy of the prepared photocatalyst and hence enhanced its visible light activity. Decolorization tests for MO dye solutions demonstrated the feasibility of the prepared photocatalysts, either as powder or as film immobilized on PP, to degrade organic contaminant in aqueous solutions under the UV as well as the visible light conditions. The prepared buoyant photocatalyst has the prospect to be used directly at the water surface with the natural sunlight as the light source for photocatalytic decontamination of organic pollutants in water or used water. Acknowledgment The financial support from the Ministry of Education Singapore, through R-288-000-040-112, the Academic Research Fund of National University of Singapore, is acknowledged. Supporting Information Available: Figure S1, XPS survey scans of blank PP granules and PP granules immobilized with TiO2 photocatalyst. This material is available free of charge via the Internet at http://pubs.acs.org. Literature Cited (1) Cen, H. W.; Li, X. J.; He, M. X.; Zheng, S. J.; Feng, M. Z. The Effect of Background Irradiation on Photocatalytic Efficiencies of TiO2 Thin Films. Chemosphere 2006, 62 (5), 810–816. (2) Hosseini, S. N.; Borghei, S. M.; Vossoughi, M.; Taghavinia, N. Immobilization of TiO2 on Perlite Granules for Photocatalytic Degradation of Phenol. Appl. Catal. B 2007, 74 (1-2), 53–62.
2898 Ind. Eng. Chem. Res., Vol. 48, No. 6, 2009 (3) Rao, N. N.; Chaturvedi, V. Photoactivity of TiO2-Coated Pebbles. Ind. Eng. Chem. Res. 2007, 46 (13), 4406–4414. (4) Denny, M. W. Air and Water: The Biology and Physics of Life’s Media; Princeton University Press: Princeton, NJ, 1993. (5) Liu, W. C. Water Column Light Attenuation Estimation to Simulate Phytoplankton Population in Tidal Estuary. EnViron. Geol. 2005, 49 (2), 280–292. (6) Lin, H. F.; Ravikrishna, R.; Valsaraj, K. T. Reusable Adsorbents for Dilute Solution Separation. 6. Batch and Continuous Reactors for The Adsorption and Degradation of 1,2-Dichlorobenzene from Dilute Wastewater Streams Using Titania as A Photocatalyst. Sep. Purif. Technol. 2002, 28 (2), 87–102. (7) Xu, C.; Killmeyer, R.; Gray, M. L.; Khan, S. U. M. Photocatalytic Effect of Carbon-Modified n-TiO2 Nanoparticles under Visible Light Illumination. Appl. Catal. B 2006, 64 (3-4), 312–317. (8) Zhu, J. F.; Chen, F.; Zhang, J. L.; Chen, H. J.; Anpo, M. Fe3+-TiO2 Photocatalysts Prepared by Combining Sol-Gel Method with Hydrothermal Treatment and Their Characterization. J. Photochem. Photobiol., A 2006, 180 (1-2), 196–204. (9) Ho, W.; Yu, J. C.; Lee, S. Synthesis of Hierarchical Nanoporous F-Doped TiO2 Spheres with Visible Light Photocatalytic Activity. Chem. Commun. 2006, (10), 1115–1117. (10) Mori, K. Photo-Functionalized Materials Using Nanoparticles. Photocatalysis 2004, 41, 750–756. (11) Zhao, W. K.; Fang, Y. L. Synthesis of Photocatalytic Activity TiO2 Thin Films at Low-Temperature. Acta Phys.—Chim. Sin. 2002, 18 (4), 368– 371. (12) Zhou, L.; Liu, C.; Zhao, W. K.; Hu, L.; Fang, Y. L. Photocatalytic Activity of Fe3+-Doped TiO2 Thin Films Prepared via Liquid-Phase Deposition. Chin. J. Catal. 2003, 24 (5), 359–363. (13) Zhou, L.; Zhao, W. Photocatalytic Activity of Anatase TiO2 Thin Films Prepared by Liquid Phase Deposition. Chin. J. Appl. Chem. 2002, 19 (10), 919-922. (14) Zhou, L.; Zhao, W.; Fang, Y. Photocatalytic Activity of Anatase TiO2 Thin Films Prepared by Direct Deposition from Aqueous Solutions. Chin. J. Semiconductors 2003, 24 (5), 94-98. (15) Portjanskaja, E.; Krichevskaya, M.; Preis, S.; Kallas, J. Photocatalytic Oxidation of Humic Substances with TiO2-Coated Glass Micro-Spheres. EnViron. Chem. Lett. 2004, 2 (3), 123–127. (16) Portjanskaja, E.; Preis, S.; Kallas, J. Aqueous Photocatalytic Oxidation of Lignin and Humic Acids with Supported TiO2. Int. J. Photoenergy 2006, 7. (17) Zaleska, A.; Hupka, J.; Wiergowski, M.; Biziuk, M. Photocatalytic Degradation of Lindane, p,p ′-DDT and Methoxychlor in An Aqueous Environment. J. Photochem. Photobiol., A 2000, 135 (2-3), 213–220. (18) Chen, S. F.; Liu, X. Q.; Liu, Y. Z.; Cao, G. Y. The Preparation of Nitrogen-Doped TiO2-xNx Photocatalyst Coated on Hollow Glass Microbeads. Appl. Surf. Sci. 2007, 253 (6), 3077–3082. (19) Luzuriaga, S.; Kovarova, J.; Fortelny, I. Degradation of Pre-Aged Polymers Exposed to Simulated Recycling: Properties and Thermal Stability. Polym. Degrad. Stab. 2006, 91 (6), 1226–1232.
(20) Frank, H. P. Polypropylene; Gordon and Breach Science Publishers: New York, 1968. (21) Dutschke, A.; Diegelmann, C.; Lobmann, P. Nucleation and Growth of TiO2 Thin Films on Modified Polystyrene Surfaces. Chem. Mater. 2003, 15 (18), 3501–3506. (22) Byrappa, K.; Masahiro, Y. Handbook of Hydrothermal Technology: A Technology for Crystal Growth and Materials Processing; Noyes Publications: Norwich, NY, 2001. (23) Liu, H. Y.; Gao, L. Codoped Rutile TiO2 as A New Photocatalyst for Visible Light Irradiation. Chem. Lett. 2004, 33 (6), 730–731. (24) Liu, H. Y.; Gao, L. (Sulfur, Nitrogen)-Codoped Rutile-Titanium Dioxide as A Visible-Light-Activated Photocatalyst. J. Am. Ceram. Soc. 2004, 87 (8), 1582–1584. (25) Zhu, J. F.; Deng, Z. G.; Chen, F.; Zhang, J. L.; Chen, H. J.; Anpo, M.; Huang, J. Z.; Zhang, L. Z. Hydrothermal Doping Method for Preparation of Cr3+-TiO2 Photocatalysts with Concentration Gradient Distribution of Cr3+. Appl. Catal. B 2006, 62 (3-4), 329–335. (26) Gole, J. L.; Stout, J. D.; Burda, C.; Lou, Y. B.; Chen, X. B. Highly Efficient Formation of Visible Light Tunable TiO2-xNx Photocatalysts and Their Transformation at The Nanoscale. J. Phys. Chem. B 2004, 108 (4), 1230–1240. (27) Yukselici, M. H. Two Different Mechanisms of Formation of Quantum Dots in Borosilicate Glass. J. Phys.: Condes. Matter 2001, 13 (27), 6123–6131. (28) Cong, Y.; Zhang, J. L.; Chen, F.; Anpo, M. Synthesis and Characterization of Nitrogen-Doped TiO2 Nanophotocatalyst with High Visible Light Activity. J. Phys. Chem. C 2007, 111 (19), 6976–6982. (29) Moulder, J. F.; Stickle, W. F.; Sobol, P. E.; Bombon, K. D. Handbook of X-ray Photoelectron Spectroscopy; Perkin-Elmer Corp.: Eden Prairie, MN, 1992. (30) Cong, Y.; Chen, F.; Zhang, J. L.; Anpo, M. Carbon and NitrogenCodoped TiO2 with High Visible Light Photocatalytic Activity. Chem. Lett. 2006, 35 (7), 800–801. (31) Huo, Y.; Bian, Z.; Zhang, X.; Jin, Y.; Zhu, J.; Li, H. Highly Active TiO2-xNx Visible Photocatalyst Prepared by N-Doping in Et3N/EtOH Fluid under Supercritical Conditions. J. Phys. Chem. C 2008, 112 (16), 6546– 6550. (32) Burda, C.; Lou, Y. B.; Chen, X. B.; Samia, A. C. S.; Stout, J.; Gole, J. L. Enhanced Nitrogen Doping in TiO2 Nanoparticles. Nano Lett. 2003, 3 (8), 1049–1051. (33) Huang, J. Sol-Gel Mechanism and Technology; Chemical Industry Press: Beijing, 2005. (34) Surendran, K. P.; Wu, A.; Vilarinho, P. M.; Ferreira, V. M. Solgel Synthesis of Low-Loss MgTiO3 Thin Films by a Non-Methoxyethanol Route. Chem. Mater. 2008, 20 (13), 4260–4267.
ReceiVed for reView September 9, 2008 ReVised manuscript receiVed December 30, 2008 Accepted January 7, 2009 IE801362A