Effect of Thickness of Photocatalyst Film Immobilized on a Buoyant

Sep 15, 2011 - Effect of Thickness of Photocatalyst Film Immobilized on a Buoyant Substrate on the Degradation of Methyl Orange Dye in Aqueous Solutio...
1 downloads 0 Views 1MB Size
ARTICLE pubs.acs.org/IECR

Effect of Thickness of Photocatalyst Film Immobilized on a Buoyant Substrate on the Degradation of Methyl Orange Dye in Aqueous Solutions under Different Light Irradiations Hui Han and Renbi Bai* Department of Civil and Environmental Engineering, Faculty of Engineering, National University of Singapore, 1 Engineering Drive 2, Singapore 117576 ABSTRACT: This study examined the effect of thickness of photocatalyst film immobilized on a buoyant polypropylene (PP) substrate on the degradation of methyl orange (MO) dye in aqueous solutions using different light sources. The photocatalyst used was titanium dioxide (TiO2) and the immobilized photocatalyst film thickness varied in a range from 6.9 to 34.1 μm. Both degradation experiments and model fitting analysis for the experimental data were carried out. All the experimental results on MO dye degradation were found to be well fitted by a pseudo-first-order reaction kinetic model. However, the degradation performance and the reaction kinetic model rate constant showed a clear trend of increase with the increase of the photocatalyst film thickness under the visible (Vis) light irradiation, but had no obvious change with the photocatalyst film thickness under the ultraviolet (UV) light irradiation. This phenomenon was attributed to the much smaller active photocatalyst film thickness for the UV light (due to the lesser quantity of photons and hence less thickness through which the light penetrated), as compared to that for the Vis light in the photocatalytic reaction. The study reveals that methods effectively increasing the photocatalyst film thickness on the buoyant substrate in the preparation of the buoyant composite photocatalyst are beneficial for the photocatalytic degradation of MO dye, especially for situations when the system may work mainly with the visible light, such as using the natural sunlight as the light source.

1. INTRODUCTION Photocatalysis has been widely studied as an effective method to remove organic contaminants, especially those that are nonbiodegradable, from water or wastewater. Titanium oxide (TiO2) in the nanoparticle form, such as Degussa P25, has been commonly used as the photocatalyst. Although photocatalysts in the nanoparticle form may have higher photocatalytic reactivity (as compared to those in larger particle sizes) and can be well dispersed into the water to be treated to form slurry reactors, a major problem related to those applications has been the difficulty and high cost of separating the nanoparticle photocatalysts from the treated water. In spite of the fact that new separation technologies such as membrane filtration may be used as an effective method to separate the photocatalyst nanoparticles from the treated water, membrane fouling is often severe, due to the nanoparticles plugging the membrane pores or forming a tight cake layer on the membrane surface, which not only causes significant reduction in the water permeate flux and increase in the operational energy consumption, but also even renders a complete loss of the membrane’s reusability.13 In an early study, we have successfully developed a buoyant composite photocatalyst with modified TiO2 film immobilized on a polypropylene substrate.4 The buoyant composite photocatalyst can float automatically to the water surface and the photocatalytic reactor was clearly divided into a top zone with the photocatalyst and a bottom zone that was entirely water (i.e., free of the photocatalyst). This development has made it convenient to separate the photocatalyst from the treated water. In addition, while many available photocatalysts were only active under the UV light, the developed buoyant composite photocatalyst showed excellent activity under both the UV and the visible r 2011 American Chemical Society

light. This latter feature provides a great prospect for the natural sunlight to be used as an effective light source in the photocatalytic reactions for water or wastewater treatment to save energy cost. The reason is that natural sunlight is well-known to contain a large portion of the visible light (about 4550%), as compared to that of the UV light (only about 35%).5 Besides, another major consideration in the development of the buoyant composite photocatalyst is to increase the light utilization efficiency. In the conventional practices where photocatalyst nanoparticles are commonly applied in a slurry form in photocatalytic reactors, no matter whether the used photocatalysts have visible light activity or not and whether the UV light or visible light (such as sunlight) is used as the light source, the light utilization efficiency for the photocatalytic reaction is usually low. This is attributed to the fact that light, especially the UV light, attenuates significantly in water with the traveling distance.6 The effect can be explained by the BeerLambert law: T ¼

I ¼ 10αl I0

ð1Þ

where T is the transmission rate of the light; I0 and I are the intensity of the light before and after penetrating the medium material, respectively; l is the light traveling distance through the medium material; and α is the attenuation coefficient of the light in the medium material. For a given light intensity, I0, supplied from the light source, the actual light intensity, I, after traveling Received: April 13, 2011 Accepted: September 15, 2011 Revised: August 16, 2011 Published: September 15, 2011 11922

dx.doi.org/10.1021/ie200787j | Ind. Eng. Chem. Res. 2011, 50, 11922–11929

Industrial & Engineering Chemistry Research through a specific distance, l, in a medium material (e.g., air or water solution, etc.) for the photocatalytic reaction will depend on the light attenuation coefficient α in the medium material. It has been reported that the attenuation coefficient α in water is at least 100 times greater than that in air.6 Therefore, it is advantageous to use the buoyant composite photocatalyst so that the photocatalytic reactions can take place at the air/water interface rather than in water, with the light lamps installed above the water surface or using the natural sunlight as the light source. In the literature, photocatalytic treatment of water or wastewater with photocatalysts in the nanoparticle form has been well reported, including the degradation efficiency and the effect of various operational parameters.711 The parameters that have been found to affect the reaction kinetics include water pH value, photocatalyst concentration, light intensity, and air flow rate, etc. For photocatalysts in the nanoparticle form, the particle sizes have also been found to be an important material character influencing the performance. In general, smaller photocatalyst particle sizes were considered to provide larger specific surface areas,12 and hence better reactivity. However, the treatment performance was also found to depend on other factors other than photocatalyst sizes, such as the organic compounds to be degraded. For example, Bekbolet and Uyguner conducted a kinetic study in natural organic matter (NOM) degradation using commercially available Degussa P25 TiO2 and Hombikat UV-100 TiO2 powder as the photocatalysts, respectively.12 They found that the smaller size (10 nm) and higher specific surface area (189 m2/g) of the Hombikat UV-100 TiO2 photocatalyst did not render a higher adsorption capacity for NOM, as compared to the larger size (2030 nm) and smaller specific surface area (55 m2/g) of the Degussa P25 TiO2 photocatalyst. The results illustrated a case where the large sized humic acid molecules were not found to have a better treatment performance with the smaller sized photocatalyst. In contrast, photocatalysts in the film form have been relatively less studied. In the literature, there are a few directly or indirectly related reports. A work by Al-Homoudi et al. compared the performance of carbon monoxide (CO) gas sensor made of anatase TiO2 film on a sapphire substrate with a thickness of 0.25 or 1 μm.13 They found that the film with a larger thickness of 1 μm showed a greater decrease in the electrical resistance at higher CO concentrations (>40 ppm). This phenomenon has been attributed to the better crystalline quality of TiO2 in the thicker film. In another report, Choi et al. prepared TiO2 film samples with thickness from 0.26 to 0.36 μm on borosilicate glass substrate and found that the absorbance to 365 nm UV light increased with the increase of the film thickness.14 Antoniou et al.15 studied the degradation of microcystin-LR with a 6.7-μm TiO2 film on a stainless steel plate and a 0.3-μm TiO2 film on a glass plate under 365 nm UV light and reported that the degradation efficiencies, after 4 h of treatment, were surprisingly comparable (in fact to be 55% removal for the 6.7 μm film and 45% removal for the 0.3 μm film, respectively). In another study, Chen et al. prepared TiO2 photocatalyst film of different thickness on a glass substrate by dip coating of 1, 2, 4, 6, 8, and 10 times (the thickness from 6 dip coatings was 0.182 μm).16 They found that the degradation performance of creatinine under a UV lamp (300400 nm and a peak at 365 nm) increased when the TiO2 film thickness increased from dip coating of 1 to 6 times, but had no further improvement when the TiO2 film thickness increased from dip coating of 6 to 10 times. The above-mentioned studies appear to raise an issue that the thickness of TiO2 films may have

ARTICLE

an effect on their performance. However, those studies were either in fact not specifically related to photocatalytic reaction and its application or only looked at photocatalyst film in a small thickness range and with UV light only. It is therefore the objective of this study to more systematically examine the effect of photocatalyst film thickness on the photocatalytic degradation performance with the buoyant composite photocatalyst we prepared through immobilizing TiO2 nanoparticles on a polypropylene (PP) substrate in a larger photocatalyst film thickness range and with different light sources. The general method reported in an early study to immobilize TiO2 on PP with different thicknesses17 was used to obtain the photocatalyst materials in this study and their photocatalytic degradation performances were experimentally investigated. For the photocatalytic performance evaluation of the films, methyl orange (MO) dye, which is a toxic and common contaminant that cannot be effectively removed by conventional treatment methods such as biological treatment, was selected. The degradation kinetics, as a fundamental tool to obtain a better understanding of the photocatalyst’s performance, was examined as well.

2. MATERIALS AND METHODS 2.1. Preparation of the Composite Photocatalyst with Different Film Thicknesses. The general method to prepare

the buoyant composite photocatalyst with different TiO2 film thicknesses was described elsewhere in detail.17 In brief, the composite photocatalyst consisted of a TiO2 film immobilized on a polypropylene fabric (PPF) as the substrate (Sefar Tetex, Multi 05-8-620-SK-T, Switzerland) and was buoyant. The buoyant composite photocatalyst that is denoted as “RA” was composed of two layers of TiO2 on the PPF substrate surface: a flower-like rutile layer at the bottom and a N-doped anatase layer on the top; while the one denoted as “A” has only the N-doped anatase TiO2 layer on the PPF substrate.17 As reported in the previous studies,4,17 the top anatase TiO2 layer formed a solid crystalline surface in both the RA and A series. Because the anatase TiO2 was doped with nitrogen, the RA and A series of the buoyant composite photocatalyst were not only active under the UV light but also absorbed the visible light and therefore were photoactive under the visible light as well. Because the surface area of the PPF substrate was fixed, the apparent thickness of the TiO2 film on the composite photocatalyst was therefore proportional to the TiO2 loading rate—the weight of TiO2 immobilized per unit area of the PPF substrate. The loading rate of the rutile TiO2 for the RA series was fixed in the study but the loading rate of the anatase TiO2 for both the RA and A series was varied in the preparation process by controlling or adjusting the concentration of the nitrogen-doped TiO2 sol and the dip-coating times. As reported in our previous study,17 the thickness or loading rate for the A series was relatively more difficult to increase, while for the RA series, it can be easier to achieve and even reach 50% greater than that of the A series. Therefore, the total photocatalyst film thicknesses on the RA and the A series in this study were not exactly the same but they did vary to cover a wide range for the study purpose. The PPF substrate used to prepare the buoyant composite photocatalysts was in a disk shape and each had a diameter of 50 mm. The TiO2 loading rates and the film thicknesses of the prepared buoyant composite photocatalyst for this study are summarized in Table 1. 11923

dx.doi.org/10.1021/ie200787j |Ind. Eng. Chem. Res. 2011, 50, 11922–11929

Industrial & Engineering Chemistry Research

ARTICLE

Table 1. Loading Rates and Photocatalyst Film Thickness of the Prepared Buoyant Composite Photocatalyst in the RA and A Series loading rate (g 3 m2)

thickness (μm)

buoyant photocatalyst

rutile

anatase

total

rutile anatase total

RA1

60

43.6

103.6

14.2

11.2

25.4

RA2

60

54.6

114.6

14.2

14.0

28.2

RA3

60

70.8

130.8

14.2

18.2

32.4

RA4

60

77.6

137.6

14.2

19.9

34.1

A1 A2

26.8 36.8

26.8 36.8

6.9 9.5

6.9 9.5

A3

52.5

52.5

13.5

13.5

A4

70.9

70.9

18.2

18.2

Figure 1. Schematic diagram of the photocatalytic reaction system: (1) air compressor; (2) air flowmeter; (3) air distributor; (4) reactor; (5) MO dye solution; (6) buoyant composite photocatalyst; (7) light beam turning assembly; and (8) xenon lamp light source system.

2.2. Photocatalyst Characterization Analysis. The loading rate of TiO2 on the PPF substrate per unit area was determined by the weighing method as described in details elsewhere.17 The thickness of the TiO2 film on a buoyant composite photocatalyst was roughly estimated by dividing the anatase and rutile TiO2 loading rate with the density of anatase and rutile TiO2 (3.89 and 4.23 g 3 cm3), respectively. The light absorbance of the prepared buoyant composite photocatalyst was measured with a UVVis spectrophotometer (Jasco V660, Japan) equipped with a j60mm integrating-sphere accessory.4,17 The surface morphology of the RA and A series of the buoyant composite photocatalyst was observed through the SEM analysis.17 2.3. Photocatalytic Degradation Experiments for MO Dye in Aqueous Solutions. The experimental setup for the photocatalytic degradation of MO dye in this study is schematically shown in Figure 1. The reactor was a 150-mL glass beaker. An air distributor (D = 20 mm) was placed at the bottom of the reactor. Air from an air compressor was supplied to the reactor through the distributor at a flow rate of 0.1 L 3 min1, which was to improve the mass transfer as well as to provide dissolved oxygen in the solution. The reactor was filled with 100 mL of the MO dye (MO, GR grade, Sinopharm) solution at an initial concentration of 15 mg 3 L1 and with an initial pH value at around 5.8. A piece of one type of the prepared buoyant composite photocatalyst as given in Table 1 was used in an experiment. The composite photocatalyst was initially saturated in a dark box with the MO

Figure 2. Light spectra of the 150-W xenon lamp and the 100-W UV lamp.

dye solution of the same MO dye concentration as in the reactor before use in the photocatalytic reaction in the reactor to avoid or minimize the effect by the initial adsorption process on the degradation kinetics (in fact there was no noticeable adsorption effect observed). A 150-W xenon lamp (Newport, USA) with an irradiation area of 69-mm diameter (providing the UV light at 48 W 3 m 2 and the visible light at 178 W 3 m2 ) was used as the light source in most of the experiments unless mentioned otherwise.4 When the photocatalytic reaction was conducted under the full wavelength of the light from the xenon lamp, it is denoted as “UV-Vis”, while that under the visible light only is denoted as “Vis”. In the latter case, a long-wave pass filter with cut-on wavelength at 400 nm was used to block the UV light. The filter was fixed in a filter holder attached to the light beam turning assembly. Analysis confirmed that the light spectra in the 400800 nm range remained the same under the UV-Vis and Vis conditions. In some cases, a 100-W UV lamp with an irradiation area of 203-mm diameter (providing UV light at about 70 W 3 m 2 with 365 nm wavelength) was used as the light source, and the corresponding reaction is denoted as UV. The light intensity and light spectra of the two types of light lamps, as measured with a portable spectrometer (Newport, OSM2-400), are shown in Figure 2. The results confirm that the UV lamp indeed mainly provided the UV light at around 365 nm and the xenon lamp produced light from 200 to 800 nm (see the enlarged inset in Figure 2). The concentration of MO dye in the solution in each experiment was monitored by taking samples at various time intervals and analyzed with a UVVis spectrophotometer (Jasco V660, Japan) at 465 nm. 17 The total degradation time was 120 min for each of the experiments. 2.4. Degradation Kinetics Modeling. The kinetics of photocatalytic degradation of organic contaminants has usually been modeled by the LangmuirHinshelwood (LH) equation as given below:9,10,12,18 R ¼ 

dC kLH Kads C ¼ dt 1 þ Kads C

ð2Þ

where R is the reaction rate (mg 3 L 1 3 min 1 ), k L‑H is the reaction rate constant (mg 3 L 1 3 min 1 ), Kads is the adsorption coefficient of the reactant on TiO 2 (L 3 mg1 ), C is the concentration of the reactant or the target contaminant (mg 3 L 1 ) in the solution at time t, and t is the reaction time (min). 11924

dx.doi.org/10.1021/ie200787j |Ind. Eng. Chem. Res. 2011, 50, 11922–11929

Industrial & Engineering Chemistry Research

ARTICLE

Figure 3. Effect of photocatalyst film thickness on MO dye degradation performance under the Vis and UVVis irradiations by the 150-W xenon lamp.

The integration of eq 2 will lead to   C0 ln þ Kads ðC0  CÞ ¼ kLH Kads t C

ð3Þ

where C0 is the initial concentration of the reactant at t = 0. If the initial concentration C0 is relatively low, eq 3 may be approximated into an expression of the pseudo-first-order reaction rate:9,10,12,18   C0 ¼ kLH Kads t ¼ kt ln ð4Þ C where k = kLHKads (min1) may be referred to as the pseudofirst-order reaction rate constant. Equation 3 indicates that ln(C0/C) versus t during the reaction may follow a straight line and the slope of the line gives the value of k.

3. RESULTS AND DISCUSSION 3.1. TiO2 Film Thickness of the Prepared Buoyant Composite Photocatalyst. As shown in Table 1, the buoyant

composite photocatalysts prepared for this study include the RA series and the A series. The RA series had a rutile TiO2 layer as the bottom layer on the PPF substrate surface and another anatase TiO2 layer as the top layer above the rutile TiO2 layer. In contrast, the A series had only an anatase TiO2 layer on the PPF substrate surface. From an early study, it has been found that the RA preparation approach was much easier to immobilize more TiO2 on the PPF substrate to increase the photocatalyst film thickness because the bottom rutile layer provided a very rough surface.17 In this study, for all the RA series, the rutile TiO2 layer

had a fixed thickness of about 14.2 μm but the anatase TiO2 layer varied in its thickness from about 11.2 μm for RA1 to 14, 18.2, and 19.9 μm for RA2, RA3, and RA4, respectively. Similarly, the A series also had varied thickness of the anatase layer from about 6.9 μm for A1 to 9.5, 13.5, and 18.2 μm for A2, A3 and A4, respectively. Therefore, the prepared buoyant composite photocatalyst in both the RA series and A series, as listed in Table 1, provided a good array of the photocatalyst film thicknesses for the present study to compare their effect on photocatalytic degradation performance for the MO dye. 3.2. Effect of Film Thicknesses on MO Dye Degradation Performance. The typical experimental degradation results with the various types of the prepared buoyant composite photocatalyst for MO dye at an initial concentration of 15 mg 3 L1 are shown in Figure 3. It can be found from the figure that the MO dye degradation rates clearly increased with the increase of the TiO2 film thickness for the A series buoyant composite photocatalyst (from A1 to A4) under both the Vis and UVVis conditions. For the RA series, the same feature can still be observed but is less apparent. To confirm this finding, the pseudo-first-order reaction rate constants, k, under the Vis and UVVis conditions, i.e., kVis and kUVVis, are determined from all the relevant experimental results as discussed in Section 2.4 and are given in Table 2. The linear regression coefficient, i.e., the R2 value, from the experimental results to obtain the corresponding k value, is also included in Table 2 for each case. The results in Table 2 reveal that the experimental degradation data fitted the pseudo-first-order kinetic reaction model very well and the pseudo-first-order reaction rate constants, k, for both the A and the RA series of the buoyant composite photocatalyst show a clear trend of increase with the increase of the photocatalyst film 11925

dx.doi.org/10.1021/ie200787j |Ind. Eng. Chem. Res. 2011, 50, 11922–11929

Industrial & Engineering Chemistry Research

ARTICLE

Table 2. Pseudo-First-Order Reaction Rate Constants Determined from the MO Dye Degradation Experiments under the Irradiation of the 150-W Xenon Lamp Vis

UVVis 1

k UVVis (min1)

buoyant photocatalyst

anatase TiO2 thickness (μm)

R

k Vis (min )

R

RA1

11.2

0.999

0.00431

0.999

0.00561

RA2

14.0

0.999

0.00445

0.993

0.00585

RA3 RA4

18.2 19.9

0.999 0.999

0.00495 0.00503

0.999

0.00641

A1

6.9

0.999

0.00264

0.998

0.00504

A2

9.5

0.999

0.00354

0.999

0.0059

A3

13.5

0.998

0.00409

0.999

0.00681

A4

18.2

0.997

0.00513

0.999

0.00787

2

2

thickness under either the Vis or UVVis irradiation condition. Thus, the experimental results and the model fitting analysis seem to indicate that increasing the photocatalyst film thickness on the prepared buoyant composite photocatalyst is always advantageous to the photocatalytic reaction in MO dye degradation. To clarify this point, the photocatalytic reaction which involves heterogeneous photocatalysts may be further examined in more detail. The reaction can be a complex process at the molecular level and possibly include a number of sequential steps. First, MO dye molecules were adsorbed on or diffused to the surface of the TiO2 film: MOdyemolecules f fMOdyemoleculegads

ð5Þ

Next, light was absorbed by the TiO2 film and the pairs of electrons and holes were generated (MO dye sensitized mechanism may be involved when under the visible light irradiation) TiO2 þ hν f e þ hþ

ð6Þ

In the meantime, hydroxyl radical OH 3 and hydroperoxyl radical OOH 3 were also generated by the reaction of the electrons and holes with the adsorbed H2O and O2 molecules. Finally, the active groups (OH 3 , OOH 3 , and h+) attacked the MO molecules and eventually oxidized and degraded them. It may be reasonable to assume that all the reaction steps except that in eq 6 would probably mainly involve the surface of the TiO2 photocatalyst film and, since the outer surface was dense,17 the buoyant composite photocatalyst in the A1 to A4 series or in the RA1 to RA4 series had a compatible outer surface area. The reaction in eq 6, however, can involve not only the surface but also the bulk of the photocatalyst film.19 The thick photocatalyst film may be able to absorb more photons from the light and thus generate more electron/hole pairs, in comparison with the thin photocatalyst film. As a result, higher concentrations of the active radicals were available, which resulted in higher reaction rates in MO dye degradation by the buoyant composite photocatalyst with thicker photocatalyst films. 3.3. Effect of the UV and Vis Light on the Performance of the Buoyant Composite Photocatalyst with Different Photocatalyst Film Thicknesses. The results in the previous section can provide affirmative support to the positive effect of increasing the photocatalyst film thickness on improving the MO dye degradation performance under the Vis condition. However, there is no clear information on the effect under the UV condition can be derived. Although the results under the UV-Vis condition showed

Figure 4. Pseudo-first-order kinetic rate constant for MO dye degradation with the A and RA series of the buoyant composite photocatalyst under the Vis and UVVis irradiations by the 150-W xenon lamp.

a similar trend to that under the Vis condition, it is not clear that the trend was caused by the visible light only or by both the visible light and the UV light together. To further explore the difference, the change in the pseudo-first-order reaction rate constant with the change of the anatase TiO2 film thickness for the A and RA series of buoyant composite photocatalyst under the UVVis or Vis irradiation is shown in Figure 4. As expected, the pseudo-firstorder reaction rate constant shows a linear increase with the increase of the photocatalyst film thickness in all the cases. The A series has a greater slope in the change than the RA series, indicating that the A series was more sensitive to the photocatalyst film thickness than the RA series. It is more interesting however to note from the figure that the line under the UVVis irradiation was parallel to and above that under the Vis irradiation for both the A and RA series of the buoyant composite photocatalyst. Since the only difference between the UVVis and Vis irradiations is that the UV light was blocked under the Vis condition, it is therefore reasonable to assume that the gap between the two lines under the UVVis and Vis conditions was caused by the effect of the UV irradiation. As the two lines under the UVVis and Vis conditions are parallel, suggesting that the gap values were constant with the increase of the photocatalyst film thickness for both the A and RA series, the results in Figure 4 therefore indicate that the MO dye degradation performance may not be changed with the increase of the photocatalyst film thickness due to the UV light irradiation. Because a light filter that can block the visible light and IR was not available, we were unable to conduct the MO dye 11926

dx.doi.org/10.1021/ie200787j |Ind. Eng. Chem. Res. 2011, 50, 11922–11929

Industrial & Engineering Chemistry Research

ARTICLE

Table 3. Pseudo-First-Order Reaction Rate Constants Determined from the MO Dye Degradation Experiments under the Irradiation of the 100-W UV Lamp UV 1 buoyant photocatalyst anatase TiO2 thickness (μm) R2 k UV (min )

RA1

11.2

0.999

0.00262

RA2

14.0

0.999

0.00268

RA3

18.2

RA4

19.9

0.998

0.00266

A1 A2

6.9 9.5

0.999 0.999

0.00288 0.0028

A3

13.5

0.999

0.00292

A4

18.2

0.999

0.00284

degradation experiments under the UV light irradiation only by using the 150-W xenon lamp that was available in the study. Therefore, another UV lamp as the light source was obtained to conduct the MO dye degradation experiments under the UV light irradiation only, even though the light intensity and spectra of the UV lamp were different from those from the 150-W xenon lamp. These experiments were to help to understand the performance of the photocatalyst with various film thicknesses under the UV light irradiation. A series of experiments similar to those with the 150-W xenon lamp before were conducted with the UV lamp and the pseudo-first-order reaction rate constant was determined for each case. The results are given in Table 3 and shown in Figure 5 for both the A and RA series of the buoyant composite photocatalyst. It is clear from Table 3 and Figure 5 that the reaction rate constants remained almost the same for all the cases with the UV irradiation (the 365 nm UV). These results therefore confirmed that the increase of the photocatalyst film thickness did not contribute to the increase in the MO dye degradation rate under the UV light irradiation. Hence, the phenomenon observed under the UVVis condition in Figure 3 was mainly caused by the visible light. Thus, it is reasonable to conclude that under the visible light irradiation, the increase of the photocatalyst film thickness on the buoyant composite photocatalyst can have a positive effect on the increase of the MO dye degradation performance. Under the UV light irradiation, however, no such effect can be observed at least for the range of photocatalyst film thickness examined in this study. 3.4. Active Photocatalyst Film Thickness under the UV and Vis Light Irradiations. To further explain the phenomenon that the photocatalytic activity increased with the increase of the photocatalyst film thickness on the buoyant composite photocatalyst under the visible light but not under the UV light, a parameter δ is defined as the active film thickness that would participate in a photocatalytic reaction and is discussed below. As reported by Cen et al.,19 a beam of 254-nm UV light with a specific intensity was completely absorbed by three layers of a TiO2 film, and an increase in the film thickness to four, six, and ten layers of the TiO2 film did not render a higher light absorbance. This result indicates that the extra layers of the TiO2 film did not participate in the light absorption. In that case, the active film thickness of the photocatalyst, δ254 for the 254-nm UV light may be considered to be the thickness of three layers of the TiO2 film. However, for the 365-nm UV light, δ365 was found to be much larger than δ254 even though the light intensity of the 254 and 365 nm light were the same. In fact, it was reported that

Figure 5. Pseudo-first-order kinetic rate constant for MO dye degradation with the A and RA series of the buoyant composite photocatalyst under the UV irradiation by the 100-W UV lamp.

even the ten layers of the TiO2 film did not completely absorb all the 365 nm light and the extent of light absorbance increased with the number of layers of the TiO2 film from one to ten layers.19 It has been shown that the active film thickness δ that can fully absorb a beam of monochromatic light is dependent on the light intensity as well as the light wavelength.19 It is also known that the two beams of 254 and 365 nm light do not have the same photon number density even though they have the same light intensity. The photon number density of a light is determined by its light energy that is up to the light wavelength and the light with shorter wavelengths has higher photon energy, as described by the following equation:20 E ¼ hv ¼ h

c λ

ð7Þ

where E is the photon energy (J); h is the Planck constant (6.626068  1034 J 3 s); v is the wavenumber (s1); c is the speed of light (equal to 299 792 458 m/s); and λ is the wavelength of the light (m). Furthermore, the photon number density of a light is determined by both the light intensity and the photon energy of the light:20 ni ¼

Ii Ii λi ¼ Ei hc

ð8Þ

where ni is the photon number density and Ii is the intensity of light with a specific wavelength i (W 3 m2). Therefore, for a beam of monochromatic light with a specific intensity Ii, the shorter the wavelength is, the higher the photon energy and the less the photon number density ni. In the case of Cen et al.’s study, the photon number density of the beam of 254-nm UV light was smaller than that of the 365-nm UV light under about the same light intensity but different photon energies. Because the band gap of the anatase TiO2 is 3.2 eV and that of the rutile TiO2 is 3.0 eV, the prepared buoyant composite photocatalyst of both the A and RA series can absorb and be excited by the 254-nm light (4.88 eV) and the 365-nm light (3.39 eV). Because there were lesser numbers of photons in the beam of 254-nm light, there were therefore lesser numbers of electron/hole pairs needed to participate in the excitation and reaction. Thus, δ254 would be smaller than δ365. From this analysis, it can be concluded that when the real photocatalyst film thickness is greater than or equal to that of the active photocatalyst film thickness, the light absorbance and hence the photocatalytic 11927

dx.doi.org/10.1021/ie200787j |Ind. Eng. Chem. Res. 2011, 50, 11922–11929

Industrial & Engineering Chemistry Research

ARTICLE

reaction rate will not increase with the further increase of the photocatalyst film thickness. Similarly, when the real photocatalyst film thickness is less than that of the active photocatalyst film thickness needed, the light absorbance or photocatalytic reaction rate will increase with the further increase of the photocatalyst film thickness. In this study, the light from the 150-W xenon lamp was polychromatic light but it can be roughly divided into the UV light (200400 nm) and the visible light (400700 nm). The total photon number density n, and that contributed from the UV light, nUV, or from the visible light, nVis, may be estimated similarly from eq 8 as n¼

i ¼ 700

i ¼ 700 I

λ

i ¼ 700

i i i ¼ ¼ ∑ Ii λi ∑ hc i ¼∑200 i ¼ 200 Ei i ¼ 200 hc

nUV ¼

nVis ¼

I

1

ð9Þ

i ¼ 400

i ¼ Ii λi ≈ ∑ hc i ¼∑200 E i ¼ 200 i

I

i ¼ 400

λUV IUV IUV ¼ hc EUV

ð10Þ

i ¼ 700

I

i ¼ 700

λVis IVis IVis ¼ hc EVis

ð11Þ

1

i ¼ Ii λi ≈ ∑ hc i ¼∑400 E i ¼ 400 i

1

where λUV and λVis are the average wavelength of the UV light and the visible light, respectively; IUV and IVis are the integral light intensity of the UV light and the visible light, respectively; and EUV and EVis are the average photon energy of the UV light and the visible light, respectively. Because the integral intensity of the UV light (IUV = 48 W/m2) was much smaller than that of the visible light (IVis = 178 W/m2) and the average photon energy of UV light is much higher than that of the visible light according to eq 7, it is easy to understand that the photon number density of the UV light from the xenon lamp was much smaller than that of the visible light from the xenon lamp. Therefore, the active photocatalyst film thickness for the UV light, δUV, can be expected to be smaller than that for the visible light, δVis. This difference can explain the observed phenomenon that the MO dye degradation rate increased with the increase of the immobilized photocatalyst film thickness on the buoyant composite photocatalyst under the Vis irradiation but not under the UV irradiation. In fact, even the smallest photocatalyst film thickness on the A1 and RA1 samples may have already reached or even exceeded the active photocatalyst film thickness needed for the UV light, δUV. So, the further increase in the photocatalyst film thickness in the A2 to A4 or RA2 to RA4 series did not cause further increase in the reaction rate. Because δVis could be much larger than δUV, there should be no difficulty to understand the fact that the increase in the photocatalyst film thickness from A1 to A4 and from RA1 to RA4 indeed increased the reaction rate under the Vis light because the greatest photocatalyst film thickness on the A4 or RA4 sample may probably still not have reached the δVis yet under the light condition used in this study.

4. CONCLUSIONS The photocatalytic degradation performance of MO dye in aqueous solutions was examined with a buoyant composite photocatalyst with different thicknesses of photocatalyst film immobilized on a PPF substrate under the irradiation of a 150-W xenon lamp. Two series of the buoyant composite photocatalyst, namely RA and A, were prepared through different immobilization configuration

but both had varied thicknesses of anatase TiO2 layer on the surfaces. In all the cases, it has been found that MO dye was effectively degraded and the reaction followed a pseudo-firstorder reaction model. However, the degradation performance showed clear increase with the increase of the photocatalyst film thickness on the buoyant composite photocatalyst under the irradiation of the visible light, but their performance remained unchanged under the irradiation of the UV light. Analysis was conducted to reveal the effect of the photocatalyst film thickness to the photocatalytic reaction performance. It is shown that a light source, according to its light intensity and wavelength range, requires a specific maximum thickness of the photocatalyst film to completely absorb the light supplied (defined as an active thickness δ). When the film thickness was greater than the active thickness δ, the reaction rate did not increase with the further increase of the film thickness; if the film thickness was smaller than the active thickness, the reaction rate increased with the increase of the film thickness. As the visible light usually has a greater wavelength and also a higher light intensity (from the 150-W xenon lamp) in this study, the MO dye degradation performance therefore increased with the photocatalyst film thicknesses under the visible light irradiation, due to the greater δ, as compared to that under the UV light irradiation. The study confirmed that it can be advantageous to increase the photocatalyst film thickness on the buoyant composite photocatalyst when it is targeted for the use mainly with the visible light, such as using natural sunlight as the light source.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected]. Tel.: +65 6516 4532. Fax: +65 6779 1635.

’ ACKNOWLEDGMENT The financial support through R-288-000-067-592 from Dowtec Private Limited, Singapore, is greatly appreciated. ’ REFERENCES (1) Lee, S. A.; Choo, K. H.; Lee, C. H.; Lee, H. I.; Hyeon, T.; Choi, W.; Kwon, H. H. Use of ultrafiltration membranes for the separation of TiO2 photocatalysts in drinking water treatment. Ind. Eng. Chem. Res. 2001, 40 (7), 1712–1719. (2) Syafei, A. D.; Lin, C. F.; Wu, C. H. Removal of natural organic matter by ultrafiltration with TiO2-coated membrane under UV irradiation. J. Colloid Interface Sci. 2008, 323, 112–119. (3) Chiemchaisri, C.; Passananon, S.; Ngo, H. H.; Vigneswaran, S. Enhanced natural organic matter removal in floating media filter coupled with microfiltration membrane for river water treatment. In 4th Conference of the Aseanian-Membrane-Society Taipei, Taiwan 2007; Elsevier Science Bv: Taipei, TAIWAN, 2007; pp 335343. (4) Han, H.; Bai, R. B. Buoyant photocatalyst with greatly enhanced visible-light activity prepared through a low temperature hydrothermal method. Ind. Eng. Chem. Res. 2009, 48 (6), 2891–2898. (5) Mori, K. Photo-Functionalized Materials Using Nanoparticles: Photocatalysis. J. Soc. Powder Technol. 2004, 41, 750–756. (6) Denny, M. W. Air and Water: The Biology and Physics of Life’s Media; Princeton University Press: Princeton, NJ, 1993; pp xviii, 341. (7) Choo, K. H.; Tao, R.; Kim, M. J. Use of a photocatalytic membrane reactor for the removal of natural organic matter in water: Effect of photoinduced desorption and ferrihydrite adsorption. J. Membr. Sci. 2008, 322 (2), 368–374. 11928

dx.doi.org/10.1021/ie200787j |Ind. Eng. Chem. Res. 2011, 50, 11922–11929

Industrial & Engineering Chemistry Research

ARTICLE

(8) Huang, X. H.; Leal, M.; Li, Q. L. Degradation of natural organic matter by TiO2 photocatalytic oxidation and its effect on fouling of lowpressure membranes. Water Res. 2008, 42 (45), 1142–1150. (9) Uyguner, C. S.; Bekbolet, M. Contribution of metal species to the heterogeneous photocatalytic degradation of natural organic matter. Int. J. Photoenergy 2007, 8. (10) Uyguner, C. S.; Bekbolet, M. Photocatalytic degradation of natural organic matter: Kinetic considerations and light intensity dependence. Int. J. Photoenergy 2004, 6 (2), 73–80. (11) Zhang, P.; Jian, L. In Ozone-Enhanced Photocatalytic Degradation of Natural Organic Matter in Water. 5th World Water Congress of the International-Water-Association, Beijing, Peoples R China, 2006; Kroiss, H., Ed.; IWA Publishing: Beijing, Peoples R China, 2006; pp 5361. (12) Bekbolet, M.; Suphandag, A. S.; Uyguner, C. S. An investigation of the photocatalytic efficiencies of TiO2 powders on the decolourisation of humic acids. J. Photochem. Photobiol., A 2002, 148 (13), 121–128. (13) Al-Homoudi, I. A.; Thakur, J. S.; Naik, R.; Auner, G. W.; Newaz, G. Anatase TiO2 films based CO gas sensor: Film thickness, substrate and temperature effects. Appl. Surf. Sci. 2007, 253 (21), 8607–8614. (14) Choi, H.; Stathatos, E.; Dionysiou, D. D. Synthesis of nanocrystalline photocatalytic TiO2 thin films and particles using sol-gel method modified with nonionic surfactants. Thin Solid Films 2006, 510 (12), 107–114. (15) Antoniou, M. G.; Nicolaou, P. A.; Shoemaker, J. A.; de la Cruz, A. A.; Dionysiou, D. D. Impact of the morphological properties of thin TiO2 photocatalytic films on the detoxification of water contaminated with the cyanotoxin, microcystin-LR. Appl. Catal., B 2009, 91 (12), 165–173. (16) Chen, Y. J.; Dionysiou, D. D. Bimodal mesoporous TiO2-P25 composite thick films with high photocatalytic activity and improved structural integrity. Appl. Catal., B 2008, 80 (12), 147–155. (17) Han, H.; Bai, R. B. Highly effective buoyant photocatalyst prepared with a novel layered-TiO2 configuration on polypropylene fabric and the degradation performance for methyl orange dye under UV-Vis and Vis lights. Sep. Purif. Technol. 2010, 73 (2), 142–150. (18) Al-Rasheed, R.; Cardin, D. J. Photocatalytic degradation of humic acid in saline waters. Part 1. Artificial seawater: influence of TiO2, temperature, pH, and air-flow. Chemosphere 2003, 51 (9), 925–933. (19) 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. (20) Saleh, B. E. A.; Teich, M. C. Fundamentals of Photonics; Wiley: New York, 1991.

11929

dx.doi.org/10.1021/ie200787j |Ind. Eng. Chem. Res. 2011, 50, 11922–11929