Mesoporous Nitrogen-Doped TiO2 for the Photocatalytic Destruction

Department of Civil and Environmental Engineering, University of Cincinnati, Cincinnati, Ohio 45221-0071, and ... water contamination occurred in a di...
0 downloads 0 Views 210KB Size
Download PDF
Environ. Sci. Technol. 2007, 41, 7530-7535

Mesoporous Nitrogen-Doped TiO2 for the Photocatalytic Destruction of the Cyanobacterial Toxin Microcystin-LR under Visible Light Irradiation H Y E O K C H O I , †,§ M A R I A G . A N T O N I O U , † MIGUEL PELAEZ,† ARMAH A. DE LA CRUZ,‡ JODY A. SHOEMAKER,‡ AND D I O N Y S I O S D . D I O N Y S I O U * ,† Department of Civil and Environmental Engineering, University of Cincinnati, Cincinnati, Ohio 45221-0071, and Office of Research and Development, U.S. Environmental Protection Agency, Cincinnati, Ohio 45268

The presence of the harmful cyanobacterial toxins in water resources worldwide drives the development of an innovative and practical water treatment technology with great urgency. This study deals with two important aspects: the fabrication of mesoporous nitrogen-doped TiO2 (N-TiO2) photocatalysts and their environmental application for the destruction of microcystin-LR (MC-LR) under visible light. In a nanotechnological sol-gel synthesis method, a nitrogen-containing surfactant (dodecylammonium chloride) was introduced as a pore templating material for tailor-designing the structural properties of TiO2 and as a nitrogen dopant for its visible light response. The resulting N-TiO2 exhibited significantly enhanced structural properties including 2-8 nm mesoporous structure (porosity 44%) and high surface area of 150 m2/g. Red shift in light absorbance up to 468 nm, 0.9 eV lower binding energy of electrons in Ti 2p state, and reduced interplanar distance of crystal lattices proved nitrogen doping in the TiO2 lattice. Due to its narrow band gap at 2.65 eV, N-TiO2 efficiently degraded MC-LR under visible spectrum above 420 nm. Acidic condition (pH 3.5) was more favorable for the adsorption and photocatalytic degradation of MC-LR on N-TiO2 due to electrostatic attraction forces between negatively charged MC-LR and +6.5 mV charged N-TiO2. Even under UV light, MC-LR was decomposed 3-4 times faster using N-TiO2 than control TiO2. The degradation pathways and reaction intermediates of MC-LR were not directly related to the energy source for TiO2 activation (UV and visible) and nature of TiO2 (neat and nitrogen-doped). This study implies a strong possibility for the in situ photocatalytic remediation of contaminated water with cyanobacterial toxins and other toxic compounds using solar light, a sustainable source of energy.

* Corresponding author phone: (513)556-0724; fax: (513)556-2599; e-mail: [email protected]. † University of Cincinnati. ‡ U.S. Environmental Protection Agency. § Current address: National Risk Management Research Laboratory, U.S. Environmental Protection Agency, Cincinnati, Ohio 45268. 7530

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 21, 2007

Introduction The increasing occurrence of cyanobacteria harmful algal blooms (cyano-HABs) in water resources worldwide is alarming the environmental and health authorities because of their ability to release toxic metabolites, cyanotoxins (1). The presence of these toxins in various aquatic environments including fresh, recreational, processed, and reclaimed water has both environmental and socioeconomic impacts (2-5). Specifically, the piscary of Atlantic salmon is an important industry in the United States and Canada, and the salmon has been reported to suffer from liver disease due to the chronic exposure to a naturally occurring hepatotoxin (2). Studies in the Mediterranean region showed that the duration of cyano-HABs in some Greek lakes lasts up to 8 months while algal blooms (which can lead to cyano-HABs) occur throughout the year because of the favorable weather conditions for their growth (4). A well-documented incident of cyanobacterial water contamination occurred in a dialysis clinic in Brazil in 1996, resulting in human fatalities (5). The aforementioned incidences propelled research efforts on the detection and treatment of cyanotoxins. Most of the studies utilized a hepatotoxin from the group of microcystins, microcystin-LR (MC-LR), due to its high toxicity and frequent appearance in cyano-HABs (1, 6-13). Many technologies, including coagulation/sedimentation, activated carbon adsorption, and membrane separation, have been tested for the treatment of MC-LR (1, 6-13). Recently, titanium dioxide (TiO2) photocatalysis, one of the most effective advanced oxidation technologies (AOTs), has demonstrated high decomposition and detoxification efficiency for cyanobacterial toxins (9-13). Accordingly, extensive studies on the effect of TiO2 loading, initial MC-LR concentration, light intensity, and other chemical additives on MCLR degradation kinetics and reaction pathways have been conducted. However, to generate the reactive oxidizing species (hydroxyl radicals), which are responsible for the destruction of the toxins, the essential requirement in TiO2based AOTs is the use of ultraviolet (UV) light irradiation with high photon energy above the high band gap of TiO2 (Eg ≈ 3.2 eV) to photoexcite the catalyst. This inhibits the utilization of solar light as a sustainable energy source for TiO2 activation because only 5% of the incoming solar energy on the earth’s surface is in the UV range. Consequently, activation of TiO2 under visible light can facilitate the development of promising processes for the remediation of contaminated water resources using solar light without complicated facilities for generating and introducing UV light. To utilize visible light for TiO2 excitation, dye-sensitized or metal ion-doped TiO2 has been developed and showed promising results for the degradation of perchlorinated compounds and nitrogen oxides (14, 15). Introduction of anionic dopants, especially nitrogen, to TiO2 also makes it possible to achieve TiO2 band gap narrowing (16-18). In general, nitrogen-doped TiO2 (N-TiO2) is synthesized through two time-consuming consecutive steps: synthesis of TiO2 and then nitrogen doping of the TiO2 from various nitrogen-containing chemicals (e.g., urea, ethylamine, and gaseous nitrogen) at high temperature. In this approach, nitrogen atoms should be limited in the exterior of the TiO2 lattice (16). In TiO2 synthesis methods, one of the ways to increase nitrogen content in the TiO2 lattice is by using titanium precursors combined with a nitrogen-containing ligand, such as Ti4+-bipyridine or Ti4+-amine complexes (19, 20). Herein, we developed a simple sol-gel method to synthesize visible light activated N-TiO2 by employing a 10.1021/es0709122 CCC: $37.00

 2007 American Chemical Society Published on Web 10/05/2007

nitrogen-containing surfactant (dodecylammonium chloride, DDAC). In our previous studies, surfactant templating strategies were introduced in the sol-gel synthesis of TiO2 to tailor-design the structural properties of TiO2 and to fabricate TiO2 nanomaterials with unique reactivities and functionalities for environmental applications (21, 22). In this current study, we utilized DDAC surfactant as a pore templating material to tailor-design the structural properties of TiO2 and as a nitrogen dopant to narrow its band gap. The synthesis of mesoporous TiO2 and in situ nitrogen doping of the TiO2 were concurrently achieved. The visible light activated mesoporous N-TiO2 was evaluated for the photocatalytic destruction of MC-LR using visible and UV light. The synthesis route, physicochemical properties, and electronic structures of N-TiO2 were investigated to clarify the mode and nature of action of N-TiO2 under visible and UV light to degrade MC-LR.

Experimental Section Synthesis of Mesoporous N-TiO2. DDAC aqueous solution was prepared by reacting a stoichiometric amount of dodecylamine (DDA, Aldrich) with HCl in water as solvent (23). In such a method, the DDAC molecules self-assemble to form micelles (24). The DDAC micellar solution was stirred vigorously, and then titanium tetraisopropoxide (TTIP, Aldrich), a titanium alkoxide precursor dissolved in isopropanol (i-PrOH), was added into the DDAC solution. Hydrolysis and condensation reactions of TTIP occurred, forming a white turbid precipitate. The molar ratio of DDA: HCl:H2O:i-PrOH:TTIP was 0.5:0.5:100:30:1, where control TiO2 was also synthesized without DDA. The precipitate was condensed during drying at 20 °C for a day, finally forming a white solid TiO2/organic composite. The composite was heat-treated in a furnace (Paragon HT-22-D, Thermcraft) to increase the crystallinity of TiO2 and remove the DDAC surfactant template. The heat-treatment temperature was increased at a ramp rate of 60 °C/h to 100 °C and maintained at this temperature for 1 h. Then the temperature was increased again to 350 °C or up to 550 °C, maintained at this temperature for 2 h except for some, specifically 5 h for 350 °C, and cooled down naturally. Finally, yellowish white N-TiO2 was obtained. We completely removed any organic and inorganic materials except N-TiO2 during the synthesis procedure since any impurities from the sol ingredients might affect the overall photocatalytic performance of TiO2. Properties and Characterization of N-TiO2. A UV-vis spectrophotometer (Shimadzu 2501 PC) mounted with an integrating sphere attachment (ISR1200) for diffuse reflectance measurement was used to investigate the optical band gap of TiO2. X-ray diffraction (XRD) analysis of TiO2 using a Kristalloflex D500 diffractometer (Siemens) was employed to study crystallographic properties of TiO2. A Tristar 3000 (Micromeritics) porosimetry analyzer was used to determine the structural properties of TiO2 including Brunauer, Emmett, and Teller (BET) surface area, pore volume, and BarrettJoyner-Halenda (BJH) pore size and distribution, based on nitrogen adsorption and desorption isotherms. For the nanolevel morphology of TiO2, a JEM-2010F (JEOL) high resolution-transmission electron microscope (HR-TEM) was used after the samples were dispersed and fixed on a carboncoated copper grid (LC200-Cu, EMS). Its elemental composition was investigated using an energy dispersive X-ray spectroscope (EDX, Oxford Isis) connected with the HR-TEM. An X-ray photoelectron spectroscope (XPS, Perkin-Elmer Model 5300) with Mg KR X-rays was to used to determine the fine elemental and electronic structure of TiO2 at a takeoff angle of 15-75° and vacuum pressure of 10-8-10-9 Torr. Charge compensation was corrected, based on the C 1s peak for hydrocarbons at 284.6 eV, and a combination of 90% Gaussian and 10% Lorenzian peak shapes was utilized for

FIGURE 1. Optical UV-visible absorption spectra of control TiO2 and N-TiO2 calcined at 350 °C. curve fitting. The particle size of agglomerated TiO2 (APS) and its point of zero charge (PZC) were measured using a Zetasizer (Malvern Instruments). Photocatalytic Degradation of Microcystin-LR. TiO2 particles dispersed in water using a sonicator (2510R-DH, Bransonic) for 2 h were transferred to borosilicate reactors containing MC-LR (Calbiochem Cat #: 475815). The solution pH was adjusted at around 3.5 without buffer by adding H2SO4 solution. The reactors were completely sealed and stirred vigorously to reduce mass transfer limitation and were covered with aluminum foil to make dark conditions for adsorption of MC-LR onto TiO2 for 2 h. The reactors were then irradiated for up to 4 h by two 15 W low-pressure mercury UV tubes (Spectronics) emitting near UV radiation with a peak at 365 nm or two 15 W fluorescent lamps (Cole-Parmer) mounted with a UV block filter (UV420, Opticology) to cut the spectral range below 420 nm. The experimental conditions were as follows: volume of solution ) 10 mL, pH ) 3.5 ( 0.1 (for a comparative study, pH ) 5.7 ( 0.2 with super quality (SQ) water), MC-LR concentration ) 5.0 ( 0.1 mg/L (5 µM), TiO2 loading ) 0.5 g/L, and temperature ) 25 ( 3 °C. A sample of 0.2 mL was taken and filtered with a syringeless 0.45 µm glass microfiber vial (L815, Whatman) containing 0.2 mL of methanol as a quenching agent to stop further reaction of hydroxyl radicals generated. The samples were equally split in 0.2 mL inserts placed in vials and analyzed with high-performance liquid chromatography (HPLC, Agilent Series 1100) for the quantification of MC-LR and mass spectrometry (MS) for the determination of reaction intermediates. For the HPLC, the injection volume to a C-18 Discovery column (Supelco) at 40 °C was 50 µL. The mobile phase in isocratic mode with a flow rate of 1 mL/min was a mixture of 0.05% trifluoroacetic acid (TFA) in water and 0.05% TFA in acetonitrile at a 60:40 ratio. MC-LR was eluted at 5.4 min and measured with a photodiode array detector at 238 nm. For the identification of the intermediates with LC/MS, a gradient method was used with a mixture of 0.1% formic acid in acetonitrile and 0.1% formic acid in water (11). A Thermo Finnigan LCQ Deca ion trap mass spectrometer was utilized for the determination of the possible structures of the reaction byproducts (MS/MS analysis).

Results and Discussion Optical Band Gap. As shown in Figure 1, the absorption spectrum shoulder of N-TiO2 calcined at 350 °C was significantly extended toward the visible light range, indicating a red-shift effect of nitrogen doping. Instead of intrinsic wavelength (λgint), the linearly extrapolated wavelength was used as the effective wavelength (λgeff) for the measurement of the effective optical band gap, Egeff (25). The λgeff of N-TiO2 was approximately 468 nm (Egeff ) 2.65 eV) while that of control TiO2 was 420 nm (Egeff ) 2.95 eV). As shown in Figure 2, the Egeff of N-TiO2 steeply increased from 2.65 to 2.80 eV upon heat treatment to 400 °C and then close to 2.89 eV after VOL. 41, NO. 21, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7531

FIGURE 2. Effective band gap energy, Egeff, and nitrogen content of control TiO2 and N-TiO2 prepared at different calcination temperatures (Egeff ) 1239.6/λgeff). Note 5 h calcination time for 350 °C while it was 2 h for others. further heat treatment, compared to the stable Egeff of control TiO2 at around 2.91 eV. This decrease in Egeff of N-TiO2 was caused by the loss of nitrogen atoms incorporated in the TiO2 lattice since nitrogen atoms tend to be replaced with oxygen atoms in the air at high temperature (26). N-TiO2, especially that prepared at 350 °C, had nitrogen content up to around 1.8% while the control TiO2 had negligible nitrogen content at below 0.2% measured by XPS analysis, most probably originating from the impurities of the ingredients used. The decrease in the nitrogen content in N-TiO2 over calcination temperature is in agreement with the increase in the Egeff of N-TiO2. This implies the effect of nitrogen doping in TiO2 on the visible light response of N-TiO2. Electronic Structure. To exclude the effect of carbon in N-TiO2 on visible light response, we traced and completely removed carbon in N-TiO2 by increasing either the heat treatment temperature or the calcination time (note Figure S1 for a representation of the survey XPS spectra). The electrons of nitrogen 2p states in TiO2 are known to contribute to band gap narrowing by mixing with those of oxygen 2p states and thus N-TiO2 not only reduces the photoexcitation energy but also facilitates the transport of photocarriers to the surface for photodecomposition (16). However, the role of nitrogen in TiO2 on its visible light response is still being debated, depending on the model applied for calculating the band structures and charge densities of N-TiO2 (27, 28). Figure 3 shows the XPS spectra of control TiO2 and N-TiO2 prepared at 350 °C for Ti 2p, O 1s, and N 1s core levels. Ti 2p3/2 core levels for control TiO2 and N-TiO2 are shown at 458.8 and 457.9 eV, respectively. The 0.9 eV shift in the binding energy of electrons toward lower energy indicates that the electronic interaction of Ti with anions in N-TiO2 is different from that in control TiO2 (27). This implies that TiO2 crystal lattices are modified with substitutional and/or interstitial N atoms (e.g., Ti-N-Ti, Ti-O-N-Ti), where the electron density around Ti atoms increases since the tendency of nitrogen (i.e., Pauling’s electronegativity 3.04) to attract the bonding electrons in a chemical bond toward itself is lower than that of oxygen (i.e., 3.44). For both samples, the oxygen 1s core level appears at 530.2 eV and thus the O environment is the same (Ti-O-Ti), suggesting substitutional N atoms (i.e., Ti-N-Ti) rather than interstitial N atoms (i.e., Ti-ON-Ti) in the TiO2 lattice. The N 1s core level for N-TiO2 shows a wide peak centered at around 398.3 eV. Three binding energy regions were curve-fitted at 400.2, 398.3, and 396.4 eV, which correspond to the molecularly adsorbed N species such as NO and NO2, O-Ti-N linkages in the TiO2 lattice, and chemically bonded N- species with TiO2 lattice, respectively (19, 29, 30). As Ma et al. discussed extensively, the position of N 1s responsible for the band gap narrowing is still controversial and the presence of Ti-O-N bonds makes the XPS data interpretation difficult (30). We believe that 7532

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 21, 2007

should be understood case by case, depending on specific synthesis condition and ingredients used. As a result, even though the presence of interstitial N atoms in TiO2 lattices cannot be ruled out, the lower binding energy of N-TiO2 is likely to result from the formation of the O-Ti-N and Ti-N structures during the substitutional doping process. This leads to N 2p states on the top of the valence band of TiO2 and narrows the overall band gap of TiO2. According to elemental mapping for Ti, O, N, and C in scanning-TEM mode (note Figure S2 for the scanning-TEM image and elemental line analysis results), the N-TiO2 was composed of mainly titanium and oxygen and partially nitrogen and carbon. Even though DDA consists of mainly carbon (77.7%) and partially nitrogen (7.5%), the nitrogen signal was quite stronger than the carbon signal that originated most probably from the TEM grid with carbon mashes. The preferential nitrogen doping to TiO2 over carbon from DDAC is due to the formation of a chemical bond between N in the amine group of DDAC and the Ti metal center in TTIP (i.e., formation of N-containing Ti precursors) during sol-gel synthesis of TiO2 (20). Crystallographic and Physicochemical Properties. Some important properties of TiO2 are summarized in Table 1. The crystal phase of all TiO2 samples was anatase. In general, high crystallinity and small crystallite size are desirable for high catalytic activity of TiO2. During calcination, materials crystallinity increased while crystallite size also increased. A clear peak broadening in the (101) plane for N-TiO2 suggested small size N-TiO2 crystallites (note Figure S3 for the XRD of TiO2). It is interesting to note that N-TiO2 exhibited smaller (101) plane dspace compared to control TiO2 (also note Figure S4 for the variation of (101) plane dspace). Due to a higher atomic radius of N atom (0.65 Å) compared to O atom (0.60 Å), substitutional N for O in TiO2 lattice results in a decrease in the interplanar distance. Once heat treatment temperature increased to 550 °C, dspace was restored to 3.518 Å, typical dspace for (101) of pure anatase due to the loss of N. N2 adsorption-desorption isotherms of N-TiO2 particles were type IV, typical for mesoporous materials (note Figure S5 for the isotherms and pore size distribution of TiO2). The BJH pore size distribution of N-TiO2 calcined at 350 °C was narrow, ranging from 2 to 8 nm. Its BET surface area and pore volume are significantly high at 150 m2/g and 0.202 cm3/g, respectively. During calcination, the overall structural properties, except for crystallinity, were deteriorated due to the collapse of initial porous structure and growth of crystallites. However, the structural properties of N-TiO2 are much better than those of control TiO2, suggesting that the DDAC surfactant effectively acted as a pore template. Nanocrystalline N-TiO2 clusters of average size of 152 nm exhibited a highly porous interconnected inorganic network, yielding high porosity of 44.0%, compared to dense control TiO2 clusters with low porosity of 25.1% and surface area of 72.5 m2/g and large aggregate size of 226 nm (note Figure S6 for the various HR-TEM images of TiO2). Formation of Mesoporous N-TiO2. The critical micelle concentration of DDAC aquatic solution at pH 4-7 is around 10-2 mol/L, above which the DDAC has been reported to form a well-defined self-assembled structure (24, 31). In this study, the DDAC concentration in water is around 2.8 × 10-1 mol/L. When the liquid molecular precursor of titanium is added to DDAC solution, TTIP is hydrolyzed and condensed around the self-assembled surfactants, forming a surfactant organic core/TiO2 inorganic shell composite, as demonstrated in Scheme 1. The effect of DDAC in the TiO2 network is clearly seen even before calcination (note Figure S6). The inorganic network of control TiO2 is a continuous phase while that of the composite exhibits a much less condensed phase. During thermal treatment, the surfactant templates are removed, leaving a porous structure. In addition, nitrogen atoms in

FIGURE 3. High-resolution XPS analysis of control TiO2 and N-TiO2 calcined at 350 °C: (a) at 456-466 eV for Ti 2p, (b) at 526-536 eV for O 1s, and (c) at 394-404 eV for N 1s.

TABLE 1. Crystallographic and Physicochemical Properties of Control TiO2 and N-TiO2 Prepared at Different Calcination Conditions crystallographicb phase

CS% (%)

D(101) (Å)

SA (m2/g)

PV (cm3/g)

PV% (%)

350 400 500

anatase anatase anatase

14.6 16.2 22.8

82 85 96

3.495 3.498 3.515

72.5 68.1 48.7

0.086 0.075 0.047

25.1 22.6 15.5

350 400 450 500 550

anatase anatase anatase anatase anatase

10.4 11.6 12.8 14.5 17.4

83 86 89 94 98

3.467 3.471 3.489 3.510 3.518

150 147 125 102 87.4

0.202 0.183 0.167 0.149 0.132

44.0 41.6 39.4 36.7 33.9

TiO2

N-TiO2

sample

physicochemicalb

CS (nm)

calcination temp.a (°C)

PS (nm)

4.5 4.8 5.7 6.9 8.4

APS (nm)

PZC

226 256 324

5.3 5.5 5.9

152 168 175 206 268

6.2 6.5 6.7

a Note 5 h calcination time for 350 °C while it was 2 h for others. b CS, crystallite size; CS%, crystallinity; D (101), dspace in (101) plane; SA, BET surface area; PV, pore volume; PV%, porosity; PS, BJH average pore size from adsorption isotherm branch; APS, agglomerated particle size; PZC, point of zero charge.

SCHEME 1. Incorporation of Ti-O-Ti Network onto Self-organized DDAC Surfactant Micelles To Form an Organic Core/Inorganic Shell Composite, Followed by the Removal of the Organic Templates To Form N-TiO2 with Mesoporous Structure

the DDAC surfactant are diffused and incorporated into the crystal lattice of TiO2 as either Ti-N-Ti or Ti-N-O-Ti environment. Due to the formation of nitrogen-containing Ti precursors as discussed previously, nitrogen atoms are already incorporated with the TiO2 nuclei consisting of several unit cells during sol-gel synthesis of TiO2 (20). When the nuclei crystallize and grow during calcination, nitrogen atoms can easily diffuse into TiO2 lattices in the whole grain. The distribution of N in TiO2 was proved by tracking the amount of nitrogen in N-TiO2 prepared at 350 °C using XPS depth profiling. Nitrogen content at X-ray incident angles at 15, 45, and 75°, corresponding to a surface depth of approximately 2-3, 4-5, and 6-7 nm, was relatively stable at 1.88 ( 0.35, 1.81 ( 0.26, and 1.51 ( 0.31%, respectively. Photocatalytic Destruction of MC-LR under Visible Light. As shown in Figure 4, no photolysis of MC-LR under visible light in the absence of TiO2 was observed. The dark adsorption of MC-LR with a molecular size of 1.2-2.6 nm on

2-8 nm mesoporous N-TiO2 occurred immediately and reached equilibrium within 2 h (1, 6). Afterward, MC-LR was effectively destroyed under visible light above 420 nm using N-TiO2 with narrowed band gap energy, Egeff ) 2.65 eV (468 nm). Acidic pH condition was favorable for both the adsorption and photocatalytic degradation of MC-LR. Based on the surface chemistry of MC-LR and TiO2 at pH 3.5, the surface of MC-LR is overall negatively charged by the dissociation of its free carboxylic groups while that of N-TiO2 with a PZC of 6.2 is positively charged at +6.5 mV, resulting in electrostatic attraction forces between MC-LR and N-TiO2 (7, 10). On the other hand, at pH 5.7, adsorption of negatively charged MC-LR onto relatively neutral N-TiO2 (+0.9 mV) was negligible. The difference in the degradation kinetics at the pH conditions is obvious, when considering that the first step of photocatalytic oxidation is adsorption of MC-LR on the TiO2 surface since the lifetime of hydroxyl radicals is very short and thus they react with mainly adsorbed species. The degradation kinetics followed pseudo-first-order reaction and VOL. 41, NO. 21, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7533

FIGURE 4. Adsorption followed by photocatalytic degradation of MC-LR using N-TiO2 prepared at 350 °C under visible light (>420 nm) at different pH conditions.

FIGURE 5. MC-LR degradation efficiency of control TiO2 and N-TiO2 at pH 3.5 after visible light (>420 nm) irradiation for 30 and 120 min. The initial concentration of MC-LR was reset as its concentration after 2 h adsorption equilibrium in dark conditions. “Control” in the x-axis represents all control TiO2 samples calcined at 350, 400, and 500 °C. the apparent reaction rate constant, k at pH 3.5 (k ) 0.026 min-1), was 2.5 times higher than that at pH 5.7 (k ) 0.010 min-1). The summarized results of MC-LR degradation under visible light are shown in Figure 5. The concentration of MCLR, after 2 h dark adsorption where approximately 10-15% of MC-LR disappeared, was reset as its initial concentration for measuring the photocatalytic degradation efficiency. MCLR degradation under visible light proved well the obvious role of N doping, band gap narrowing, and thus visible light response of TiO2. MC-LR degradation using control TiO2 and even P25 (well-defined 30 nm TiO2 nanoparticles, Degussa, Germany), which is a benchmark TiO2 photocatalyst with high activity, was negligible because of their higher Egeff than the photon energy provided by the visible light. As expected from its Egeff shown in Figure 2, N-TiO2 calcined at 350 °C could destroy 50% of MC-LR within 30 min and almost completely within 2 h and the other N-TiO2 photocatalysts also exhibited visible light activation for the degradation of MC-LR. However, a further increase in calcination temperature resulted in a significant decrease in MC-LR degradation efficiency of N-TiO2. As a result, the activity of TiO2 under visible light is in agreement with mainly its nitrogen content demonstrated in Figure 2, rather than any other properties including crystallinity and surface area. Photocatalytic Destruction of MC-LR under UV Light. No photolysis of MC-LR even under UV-365 nm irradiation was observed since the maximum light absorbance of MCLR is at 238 nm. As shown in Figure 6, the advantage of utilizing mesoporous N-TiO2 with high surface area in degrading MC-LR under UV light was studied in comparison with that of control TiO2. As expected, both control TiO2 and N-TiO2 effectively destroyed MC-LR and the MC-LR degradation kinetics under UV light was much faster than that under visible light. In the case of N-TiO2, MC-LR was almost completely decomposed within 30 min. The apparent first7534

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 21, 2007

FIGURE 6. MC-LR degradation by control TiO2 and N-TiO2 at pH 3.5 under UV-365 nm irradiation. The initial concentration of MC-LR was reset as its concentration after 2 h adsorption equilibrium in dark conditions. order reaction rate constants of MC-LR using N-TiO2 prepared at 350 and 400 °C were 0.064 and 0.088 min-1, respectively, which are 3-4 times higher than those using the control TiO2. As demonstrated in Table 1, this is due to much enhanced structural properties of N-TiO2 (i.e., high surface are, small crystallite size, and less agglomeration). Based on the results, the activity of TiO2 under UV light seemed to be dependent on mainly its structural properties since no significant difference in UV-365 nm light absorbance between N-TiO2 and control TiO2 was observed in Figure 1. Increase in the MC-LR decomposition rate by TiO2 calcined at elevated temperature can be ascribed to the slight increase in crystallinity from 82% to 85% for control TiO2 and from 83% to 86% for N-TiO2. Short Discussion on Comparative Degradation Pathways. We investigated comparative photocatalytic degradation pathways of MC-LR using N-TiO2 under visible and UV light (note Figure S7 for a representation of the formation of reaction intermediates and short description). The reaction intermediates found under UV light activation were observed under visible light activation as well but in lower ion counts due to the slower activation of N-TiO2 under visible light. No new intermediates were found so far, compared to those reported elsewhere employing P-25 TiO2 slurry systems under UV light, and therefore we will not discuss our findings in detail here (11, 32). In our study, all the proposed structures of these reaction byproducts are supported by the MS/MS analysis (data not shown here). These results suggest that the degradation pathways and reaction intermediates of MCLR depend on the reaction kinetics and treatment times, irrespective of the energy source for TiO2 activation (UV and visible) or nature of TiO2 (neat and nitrogen-doped). It should be noted that the intermediates we found are not toxic anymore since alternations in the Adda chain of MC-LR prohibit the toxin from binding properly in the receptor of the protein phosphatases (note Figure S7 for the removal of the chain of the Adda moiety) (32, 33). For the design of solar-driven treatment technologies, this study implies a strong possibility for the in situ photocatalytic remediation of contaminated water with cyanobacterial toxins and other toxic organic compounds using solar light, a sustainable source of energy.

Acknowledgments This research was funded in part by the National Science Foundation through a CAREER award (BES-0448117) to Dionysios D. Dionysiou, the U.S. Environmental Protection Agency (RD-83322301), and the Center of Sustainable Urban Engineering (SUE) at the University of Cincinnati. M.G.A. is grateful to Sigma Xi Scientific Society for a Grant-in-Aid of Research Fellowship.

Supporting Information Available Survey XPS spectrum (Figure S1), scanning-TEM image and line analysis results (Figure S2), XRD patterns (Figure S3), variation of (101) plane dspace (Figure S4), N2 adsorptiondesorption isotherms and pore size distribution (Figure S5), HR-TEM morphology (Figure S6), and reaction intermediates (Figure S7). This material is available free of charge via the Internet at http://pubs.acs.org.

(16)

Literature Cited

(19)

(1) Antoniou, M. G.; de la Cruz, A. A.; Dionysiou, D. D. Cyanotoxins: New generation of water contaminants. J. Environ. Eng. 2005, 131, 1239. (2) Andersen, R. J.; Luu, H. A.; Chen, D. Z. X.; Holmes, C. F. B.; Kent, M. L.; Le Blanc, M.; Taylor, F. J. R.; Williams, D. E. Chemical and biological evidence links microcystins to salmon ‘netpen liver disease’. Toxicon 1993, 31, 1315. (3) Mohamed, Z. A.; Carmichael, W. W.; Hussein, A. A. Estimation of microcystins in the freshwater fish Oreochromis niloticus in an Egyptian fish farm containing a Microcystis bloom. Environ. Toxicol. 2003, 18, 137. (4) Cook, C. M.; Vardaka, E.; Lanaras, T. Toxic cyanobacteria in Greek freshwaters, 1987-2000: Occurrence, toxicity, and impacts in the Mediterranean region. Acta Hydrochim. Hydrobiol. 2004, 32, 107. (5) Pouria, S.; de Andrade, A.; Barbosa, J.; Cavalcanti, R. L.; Barreto, V. T. S.; Ward, C. J.; Preiser, W.; Poon, G. K.; Neild, G. H.; Codd, G. A. Fatal microcystin intoxication in haemodialysisunit in Caruaru, Brazil. Lancet 1998, 352, 21. (6) Lawton, L. A.; Robertson, P. K. J. Physicochemical treatment methods for the removal of microcystins (cyanobacterial hepatoxins) from potable water. Chem. Soc. Rev. 1999, 28, 217. (7) Song, W.; de la Cruz, A.; Rein, K.; O’shea, K. E. Ultrasonically induced degradation of microcystin-LR and -RR: Identification of products, effect of pH, formation and destruction of peroxides. Environ. Sci. Technol. 2006, 40, 3941. (8) Lee, J.; Walker, H. W. Effect of process variables and natural organic matter on removal of microcystin-LR by PAC-UF. Environ. Sci. Technol. 2006, 40, 7336. (9) Lawton, L. A.; Robertson, P. K. J.; Cornish, B. J. P. A.; Jaspers, M. Detoxification of microcystins (cyanobacterial toxins) using TiO2 photocatalytic oxidation. Environ. Sci. Technol. 1999, 33, 771. (10) Feitz, A. J.; Waite, D. T.; Jones, G. J.; Boyfen, B. H.; Orr, P. T. Photocatalytic degradation of the blue green algal toxin microcystin-LR in a natural organic aqueous matrix. Environ. Sci. Technol. 1999, 33, 243. (11) Lui, I.; Lawton, L. A.; Robertson, P. K. J. Mechanistic studies of the photocatalytic oxidation of microcystin-LR: An investigation of byproducts of the decomposition process. Environ. Sci. Technol. 2003, 37, 3214. (12) Lawton, L. A.; Robertson, P. K. J.; Cornish, B. J. P. A.; Marr, I. L.; Jaspars, M. Process influencing surface interaction and photocatalytic destruction of microcystins on titanium dioxide photocatalysts. J. Catal. 2003, 213, 109. (13) Yuan, B.; Li, Y.; Huang, X.; Liu, H.; Qu, J. Fe(VI)-assisted photocatalytic degradation of microcystin-LR using titanium dioxide. J. Photochem. Photobiol. A 2006, 178, 106. (14) Bae, E.; Choi, W. Highly enhanced photoreductive degradation of perchlorinated compounds on dye-sensitized metal/TiO2 under visible light. Environ. Sci. Technol. 2003, 37, 147. (15) Lin, Y.-M.; Tseng, Y.-H.; Huang, J.-H.; Chao, C. C.; Chen, C.-C.; Wang, I. Photocatalytic activity for degradation of nitrogen

(17) (18)

(20) (21)

(22)

(23) (24) (25) (26)

(27) (28) (29) (30) (31) (32) (33)

oxides over visible light responsive titania-based photocatalysts. Environ. Sci. Technol. 2006, 40, 1616. Asahi, R.; Morikawa, T.; Ohwaki, T.; Aoki, K.; Taga, Y. Visiblelight photocatalysis in nitrogen-doped titanium oxides. Science 2001, 293, 269. Irie, H.; Wanatabe, Y.; Hashimoto, K. Nitrogen-concentration dependence on photocatalytic activity of TiO2-xNx powders. J. Phys. Chem. B 2003, 107, 5483. Burda, C.; Lou, Y.; Chen, X.; Samia, A. C. S.; Stout, J.; Gole, J. L. Enhanced nitrogen doping in TiO2 nanoparticles. Nano Lett. 2003, 3, 1049. Sano, T.; Negishi, N.; Koike, K.; Takeuchi, K.; Matsuzawa, S. Preparation of a visible light-response photocatalyst from a complex of Ti4+ with a nitrogen-containing ligand. J. Mater. Chem. 2004, 14, 380. Belver, C.; Bellod, R.; Fuerte, A.; Fernandez-Garcia, M. Nitrogencontaining TiO2 photocatalysts: Part 1. Synthesis and solid characterization. Appl. Catal. B 2006, 65, 301. Choi, H.; Sofranko, A. C.; Dionysiou, D. D. Nanocrystalline TiO2 photocatalytic membranes with a hierarchical mesoporous multiplayer structure: Synthesis, characterization, and multifunction. Adv. Funct. Mater. 2006, 16, 1067. Choi, H.; Stathatos, E.; Dionysiou, D. D. Sol-gel preparation of mesoporous photocatalytic TiO2 films and TiO2/Al2O3 composite membranes for environmental applications. Appl. Catal. B 2006, 63, 60. Loh, W.; Teixeira, L. A. C.; Lee, L.-T. Isothermal calorimetric investigation of the interaction of poly(N-isopropylacrylamide) and ionic surfactants. J. Phys. Chem. B 2004, 108, 3196. Dai, Q.; Laskowski, J. S. The Krafft point of dodecylammonium chloride: pH effect. Langmuir 1991, 7, 1361. Sun, B.; Smirniotis, P. G.; Boolchand, P. Visible light photocatalysis with platinized rutile TiO2 for aqueous organic oxidation. Langmuir 2005, 21, 11397. Sathish, M; Viswanathan, B.; Viswanath, R. P.; Chinnakonda, S. G. Synthesis, characterization, electronic Structure, and photocatalytic activity of nitrogen-doped TiO2 nanocatalyst. Chem. Mater. 2005, 17, 6349. Saha, N. C.; Tompkins, H. G. Titanium nitride oxidation chemistry: An X-ray photoelectron spectroscopy study. J. Appl. Phys. 1992, 72, 3072. Xu, T.-H.; Song, C.-L.; Liu, Y.; Han, G.-R. Band structures of TiO2 doped with N, C and B. J. Zhejiang Univ. Sci. 2006, 7, 299. Wagner, C. D.; Riggs, W. M.; Davis, L. E.; Moulder, J. F.; Muilenberg, G. E. Handbook of X-Ray Photoelectron Spectroscopy; Perkin-Elmer Corporation: Eden Prairie, MN, 1979. Ma, T.; Akiyama, M.; Abe, E.; Imai, I. High-efficiency dyesensitized solar cell based on a nitrogen-doped nanostructured titania electrode. Nano Lett. 2005, 5, 2543. Aksoy, B. S. Hydrophobic Forces in Free Thin Films of Water in the Presence and Absence of Surfactants. Ph.D. Thesis. Virginia Polytechnic Institute and State University, 1997. Liu, I.; Lawton, L. A.; Cornish, B.; Robertson, P. K. J. Mechanistic and toxicity studies of the photocatalytic oxidation of microcystin-LR. J. Photochem. Photobiol. A 2002, 148, 349. Goldberg, J.; Huang, H. B.; Kwon, Y. G.; Greengard, P.; Nairn, A. C.; Kuriyan, J. Three-dimensional structure of the catalytic subunit of protein serine/threonine phosphatase-1. Nature 1995, 376, 745.

Received for review April 17, 2007. Revised manuscript received July 4, 2007. Accepted August 22, 2007. ES0709122

VOL. 41, NO. 21, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7535