Photocatalytic Degradation of 1 , 2-dichloroethane by V / TiO 2 : The Mechanism of Photocatalytic Reaction and Byproduct

The nano-sized V/TiO2 prepared by a sol-gel method calcined at 500°C for 3 h was characterized in this study. The synthesized photo-catalysts were used for the photo-degradation of 1, 2-dichloroethane (1, 2-DCE). From X-ray powder diffraction data, the crystal phase presents the mixture of anatase and rutile phases with anatase the dominant phase. From TEM images, the crystallites of photo-catalysts are spherical particles with a crystallite size about 10–20 nm. UV-Visible absorption spectra of V/TiO2 show a slightly increase in absorbance in the visible light region with the increasing vanadium ion doping concentration. The X-ray photoelectron spectroscopy results indicate that the doped vanadium ions exist in the form of V (IV). The V (IV) may alleviate the surface poison phenomenon and act as both h/e traps to reduce the recombination rate of h/e pairs, and the optimum vanadium doping amount is 0.01 mol%. The products and byproducts of 1, 2-DCE photo-degradation include H2O, CO, CO2, C2H5Cl, CH2Cl2, and HCl.


INTRODUCTION
Photo-catalytic oxidation of aqueous and gaseous contaminants has been extensively studied to solve environmental problems.It has drawn considerable academic interest as a very attractive, non-selective, room-temperature process for the degradation of a variety of organic pollutants (Fox and Dulay, 1993;Hoffmann et al., 1995).Semiconductors can provide light-induced charges for redox processes, which is primarily due to their electronic formation.In particular they are characterized by a filled valence band and empty conduction band.The elementary mechanisms of photo-catalytic transformation include a number of steps, which have been described in previous reports (Linsebigler et al., 1995).In contrast with other semiconductors, TiO 2 is widely used for environmental applications because of its physical and chemical stability, lower cost, non-toxicity, and resistance to corrosion.To initiate redox reactions, the TiO 2 surface is irradiated with light of sufficient energy to overcome the band gap.Electron-hole pairs are subsequently generated that may initiate redox reactions on TiO 2 surface.However, the absorption wavelength of TiO 2 does not conform to the solar spectrum region; the useful solar energy by the UV spectrum is less than 5%, which limits the commercial potential because of the low photoreaction rates.Further, the photo-catalytic activity is limited by the fast charge-carrier recombination and low interfacial transfer rates of charge carriers (Martin et al., 1994a;Dawson and Kamat, 2001).In order to overcome the drawbacks, some previous studies have been performed to extend the photo-response and enhance photo-catalytic activity by surface modifications of photo-catalysts.Metal ion doping is one of these surface modification technologies.Deposition of noble metals such Pt, Au, Pd, and Ru on a semiconductor surface has been demonstrated to be beneficial for photo-catalytic reactions (Dawson and Kamat, 2001;Sakthivel et al., 2004).Several transitional metal ion doping processes have been used to increase the photo-catalytic activity (Arañ et al., 2003;Sonawane et al., 2004).Some investigators reported that doping with metal ions such as Fe 3+ , Mo 5+ , V 4+ , and Rh 3+ enhanced the photo-catalytic activity, while doping with Co 3+ , Al 3+ (Choi et al., 1994), and V 5+ (Martin et al., 1994b) promoted charge-carrier recombination and inhibited the photo-catalytic activity.However, these studies mainly paid attention to aqueous contaminants, and some cases were not successful.
1, 2-dichloroethane (1, 2-DCE) is a clear, colorless, flammable, and oily liquid.The most common use of 1, 2-DCE is for the production of vinyl chloride, which is used to make a variety of plastic and vinyl products.1, 2-DCE can evaporate into the air easily from soil and water, and stays in the atmosphere for more than 5 months before it is decomposed.The International Agency for Research on Cancer and U.S. EPA have classified that 1, 2-DCE is a probable human carcinogen (USDHHS, 2001).Consequently, 1, 2-DCE was used as a model gaseous contaminant for the photo-catalytic activity test in this study.
The results of vanadium effect on photo-catalytic activity in previous studies were incompatible (Choi et al., 1994;Martin et al., 1994b).Marth et al. (1994) investigated the crystal phase transformation of 1% V/TiO 2 under various temperatures.The XRD results indicate the phase transformation temperature range of rutile from anatase is shown at 400 to 600°C.Compared to their study, the phase transformation from the anatase to rutile phase is in the higher temperatures in our study.The result is similar to our previous researches, the anatase-rutile transition occurred in the range 580-800°C while doping with metal/non-metal species (Hung et al., 2007;Lin et al., 2011).In the case of chlorine containing VOCs for the study of Choi et al. (1994) and this study, the photo-degradation efficiencies differ because the VOCs are under different phases such as gas or liquid.Therefore, it was attempted to investigate the potential mechanisms responsible for the enhanced or inhibited V/TiO 2 action on the photo-degradation of gaseous 1, 2-DCE in this study.A continuous-flow reactor was used to evaluate parameters affecting photo-catalytic activity including retention time (RT) and V dopant concentration.In addition to the traditional characterization instruments, such as X-ray powder diffraction (XRD), scanning electron microscopy (SEM), BET surface area, and UV-Visible spectrometry, other instruments, such as X-ray photoelectron spectroscopy (XPS), thermo-gravimetric/ differential thermal analysis (TG/DTA), transmission electron microscopy (TEM), and energy dispersive spectroscopy (EDS), were employed to characterize modified photocatalysts.Lastly, the byproducts of 1, 2-DCE photodegradation are qualitatively identified via fourier transform infrared spectroscopy (FTIR).
In the preparation process of Ti-precursor sol, 30 mL TTIP was mixed with 15 mL absolute ethanol.Then the mixture was slowly added into a de-ionized water-ethanol solution (water: absolute ethanol = 12:1 v/v).After stirring vigorously for 1 min, 1.2 mL HNO 3 was slowly added into the mixture.The final solution was stirred vigorously for 24 h until the Ti-precursor sol formed.The Ti-precursor sol was then dehydrated and the alcohol removed in a rotary vacuum evaporator (EYELA, N-1000S) in a 75°C water batch.In the preparation of V/TiO 2 , the predetermined amounts of vanadium (IV) oxide sulfate pentahydrate were added in the de-ionized water first, and the subsequent processes were the same as the method described above.

Photo-Catalyst Characterization
Simultaneous TG/DTA data were recorded using a TG/DTA instrument (Pyris Diamond TG/DTA, Perkin Elmer).The temperature ranged from room temperature to 1100°C in order to obtain crystallization and phasetransformation data.All analyses were performed in a flowing air atmosphere with the heating rate of 10 °C/min.The crystalline structures of photo-catalysts were determined by XRD (D/MAX III-V, RIGAKU) using CuKα radiation with a nickel filter.The applied current and voltage were 30 mA and 40 kV, respectively.The scan ranged from 10 to 80 (2 theta degree) with a scan rate of 6° min -1 .All peak data measured by XRD analysis were assigned to known crystalline phases by comparing with those of JCPDS database.The surface area was measured with a Micromeritics ASAP 2010 instrument using adsorption of nitrogen at 77 K. Prior to adsorption measurements, the samples were degassed under a vacuum of 7 × 10 -4 kPa at 110°C for 3 h.The surface area was calculated by BET equation.The surface morphology was observed with a SEM instrument (S3000-N, HITACHI).In order to prevent the charge build-up during SEM observations, samples were coated with gold.Further, TEM images of nano-scale samples were obtained on a TEM instrument (CM-200 TWIN, PHILIPS).The samples were ultrasonically dispersed in acetone to form dilute suspensions (1.3 mg/mL) and then dropped a few droplets on copper grids with carbon film coated (200 mesh, TED PELLA).After that the copper grids were put in a dry cabinet and dried at room temperature before loading into the TEM.UV-Visible spectra of the photo-catalysts were recorded using a UV-Vis spectrophotometer (Lamda 35, Perkin Elmer) equipped with an integrating sphere.A BaSO 4 disc was used as a reference.The scan ranged from 200 nm to 900 nm.All spectra were monitored in the absorption mode and acquired under ambient conditions.To analyze the binding energy of the samples, XPS were performed by an ESCA commercial instrument (Sigma Probe, Thermo VG-Scientific) using a source of Al Kα-radiation with the energy of 1486.6 eV.The binding energies were calibrated with reference to C 1s at 285 eV.

Photo-Catalytic Degradation
The photo-catalytic degradation processes were carried out in a continuous flow reactor at room temperature.The cylinder reactor was made of pyrex glass (6.0 cm i.d., 6.4 cm o.d., and 15.5 cm length), quartz light window (2.5 cm i.d., 3.0 cm o.d., and 60 cm length), and stainless steel assembly.The appropriate amounts of photo-catalysts were dissolved in a water-ethanol solution (water: absolute ethanol = 3:7 v/v) to form the coating solution.Pyrex glass was used as the substrate for coating catalysts via spincoating method.A 6 W black-light UV lamp (BLE-8T365, Spectronics, emission range between 310 amd 385 nm with λ max = 365 nm) was employed as a UV light source.The UV light intensity was recorded using a digital radiometer (DRC-100X, SPECTROLINE) equipped with a UV-A sensor (DIX-365A, SPECTROLINE).The 1, 2-DCE contaminated stream was generated using a programmable syringe pump (Model 220, KD-Scientific) at the preset injection rate.Inlet and outlet concentration of 1, 2-DCE was measured with an on-line gas chromatograph (5890, HEWLETT PACKARD) equipped with a flame ionization detector and fitted with a DB-624 capillary column.A sixport sampler with a 0.5 mL sampling loop was used for sampling the inlet and outlet concentrations of 1, 2-DCE.The intermediates and products of 1, 2-DCE photodegradation were analyzed using an on-line FTIR (Spectrum One, Perkin Elmer).

Thermal Analysis of V/TiO 2
TG/DTA profiles of V/Ti = 1.0 and 0.1 mol% V/TiO 2 are shown in Fig. 1.In TGA curve (Fig. 1(a)), two considerable weight losses (11% and 8%) are observed in the temperature ranges between 30 and 180°C and between 200 and 400°C, respectively.The weight loss for the latter temperature range is due to the combustion of residual organic species, including the dehydroxylation of the gel and decomposition of nitrate ions.The result is consistent with a previous study, in which the decomposition of nitrate ions was observed at around 230-400°C (Sivakumar et al., 2002).In the DTA curve (Fig. 1(a)), a broad endothermic peak centered at 110°C and extending to 200°C is attributed to dehydration and loss of water.The characteristic results are similar to those found in a previous study (Musić et al., 1997).The peaks in the 200-400°C temperature range consists of the interaction of at least two overlapping peaks.A small endothermic peak at around 200-330°C is due to the decomposition of nitrate ions.An exothermic peak at around 330-400°C and extending to 450°C may be attributed to the combustion of residual organic species and the phase transformation of amorphous to anatase phase according to the XRD pattern shown in Fig. 2. A broad exothermic peak at around 460-540°C is attributed to the phase transformation from the amorphous to anatase phase.The crystallization temperature of anatase phase in this study is similar to that of other investigations (Harizanov et al., 2001;Yu et al., 2001;Sonawane and Dongare, 2006).A board exothermic peak at around 580-680°C and extending to 750°C is attributed to the phase transformation from the anatase to rutile phase, similar to those reported by studies (Chhor et al., 1992;Qin et al., 1994).The presence of metallic ions may facilitate the transformation of the anatase to rutile by lowering the transition temperature from anatase to rutile (Pal et al., 1999).Further, vanadium-doped TiO 2 are reported to be present in the form of a solid state with a nonstoichiometric formula of V x Ti 1-x O 2 between 600 to 800°C (Martin et al., 1994b).A small exothermic peak and DTG profile (Fig. 1(a)) at around 760-800°C may be attributed to the formation of vanadium-titanium oxide.
The TG/DTA profiles (< 400°C) of V/Ti = 0.1 mol% (Fig. 1(b)) are similar to those of V/Ti = 1.0 mol%.In contrast with the TG/DTA profiles V/Ti = 1.0 mol%, the transition temperature ranges of amorphous to anatase or anatase to rutile are lower than those of V/Ti = 1.0 mol%.Further, the DTG profile of V/Ti = 0.1 mol% indicates no formation of vanadium-titanium oxide form 600 to 800°C.Consequently, the adding amounts of vanadium ions affect the thermal properties of V/TiO 2 .

X-ray Diffraction Patterns of V/TiO 2
Fig. 2 shows the XRD patterns of V/TiO 2 with V/Ti = 0.001, 0.01, 0.1, and 1.0 mol% calcined at 500°C for 3 h.The XRD patterns of V/TiO 2 show that the crystal phase presents the mixture of anatase, as the dominant phase, and rutile.Comparing with six-coordination V 5+ (0.59 Å), it is believed that V 4+ with radius of 0.63 Å can also penetrate into the TiO 2 structure (Ti 4+ radius is 0.68 Å) for a coordination number of six (Shah et al., 2002).VO 2 and TiO 2 have the same tetragonal crystal structure, but V 2 O 5 has an orthorhombic crystal structure.When the V 4+ ions present on the surface of TiO 2 , the V 4+ can penetrate into the titania lattice because of its smaller size.A broader and slight shift in the XRD patterns of anatase (1 0 1) in the range between 24.5 and 26.0° indicates that the V dopant may disorder the crystalline structure of TiO 2 .On the other hand, vanadium oxide played a role in the formation of V 2 O 5 on the TiO 2 surface between 200 and 400°C, and decomposed to form V 2 O 4 at higher temperatures (Martin et al., 1994b).Nevertheless, the XRD patterns of VO 2 , V 2 O 3 , V 2 O 4 , V 2 O 5 , (Ti 0.99 V 0.01 ) 2 O 3 , and (Ti 0.995 V 0.005 ) 2 O 3 are not shown in Fig. 2. The XRD patterns show similar phases between the nude TiO 2 and V-doped TiO 2 because of a small amount of V used.A previous study reported that coverages of at least three supported monolayers of TiO 2 are necessary before resolved XRD spectra are obtained.At lower vanadium ions added, it is not unexpected that the crystalline structure is not detected in the XRD data, especially because the photo-catalysts are composed of only one or two monolayers (Klosek and Raftery, 2001).Our observations agree with the earlier results, because the amount of vanadium doped concentration in this study is lower and thus it forms a separate phase below the detection limit of XRD.
The anatase crystallite sizes of V/TiO 2 were estimated using the Scherrer equation (Demeestere et al., 2005): where k is equal to 0.9, a shape factor for spherical particles, λ is the X-ray wavelength (λ = 0.154 nm), θ is the Bragg angle, and β is the full width at half maximum (FWHM) of the peak.The crystalline sizes are shown in Table 1.With the increase of V doping concentration, the crystallite size tends to decrease slightly compared to those of the nude TiO 2 .The crystallite sizes are between 18 and 25 nm.Further, the anatase over rutile ratio was calculated using the equation (Demeestere et al., 2005): Rutile content R (wt.%) = 100 -A (wt.%) where I A and I R are the diffraction intensities of anatase (1 0 1) and rutile (1 1 0) crystalline phases at 2θ = 25.4 and 27.5° respectively.Pal et al. (1999) pointed out that the presence of Fe may lower the transition temperature from anatase to rutile.When Fe doping amount increased, the rutile phase also increases.Further, Ruiz et al. (2005) reported that the presence of metallic ions (Ag, Pt, Cu, Co) may facilitate the transformation of the anatase to rutile.However, with the increasing of V doping concentration, the anatase over rutile ratio tends to increase slightly compared to that of the nude TiO 2 in this study.In addition, the anatase over rutile ratio of V/Ti = 0.01 mol% (80% anatase, 20% rutile) is the same as that of commercial Degussa P25 (80% anatase, 20% rutile).The changes of anatase over rutile raito upon V 4+ doping may partly explain enhanced photocatalytic activity with V-doped photocatalyst.Consequently, the adding amounts of vanadium ions affect the physical properties of V/TiO 2 .

Surface Morphologies of V/TiO 2
The surface morphologies of V/TiO 2 are shown in Fig. 3.The TEM photographs show that V/TiO 2 is indeed in the nanometer scale.Clearly, separate nano-particles of photocatalysts can be seen (ca.10-20 nm).Unfortunately, the particles were also found to aggregate during either synthesized process or pre-treatment of TEM samples.The TEM image result is consistent with the crystallite size calculated by the Scherrer equation from the XRD data.

UV-Vis Absorption Spectra of V/TiO 2
The UV-Vis absorption spectra have been examined for TiO 2 doped with various amounts of vanadium ions.Their wavelength and band gap energy are summarized in Table 1.A sudden increase of the wavelength at < 400 nm can be assigned to the intrinsic band gap absorption of TiO 2 .The UV light absorbance with the V amount (0.01 mol%) exhibits the highest UV absorbance.The increased UV absorbance may play an important role on h + /e -traps, hence enhancing the overall 1, 2-DCE conversion as discussed later.Further, red shift associated with the presence of vanadium ions can be attributed to a charge transfer transition between the vanadium ions electrons and the TiO 2 conduction or valence band.It is believed that this photo-excited electron enables chemical reactions to occur at the surface of the V-doped catalysts in the presence of visible light (Klosek and Raftery, 2001).However, since the vanadium doping concentration is extremely low (V 0.001-0.1 mol%), its effect on visible light region is not clear in this study.The band gaps of V/TiO 2 were estimated by the method described in a previous study (Yoneyama et al., 1989).

XPS Spectra of V/TiO 2
Fig. 4 shows the XPS spectra of TiO 2 and V/TiO 2 with 1 mol% calcined at 500°C for 3 h.Photoelectron peaks of Ti 2p and O 1s are clearly observed for all samples.Weak signals of V 2p are detected from V/TiO 2 , due to the low doping level, as shown in Fig. 4(b).The Ti 2p 1/2 spinorbital splitting photoelectrons for V/TiO 2 are located at binding energies of 458.6 eV and 464.3 eV, respectively, which is consistent with the values of Ti 4+ in the TiO 2 lattice (Kang and Lee, 2005).The comparison between the Ti 2p 3/2 spectra of un-doped TiO 2 and those of V/TiO 2 indicates that the latter has small binding energy shifts.It infers that the doped V ions may diffuse into TiO 2 lattices to form the V-O-Ti bond.The photoelectron spectra of O 1s can be decomposed into two peaks.The O 1s main peak at 529.9 eV is assigned to the metallic oxides, which is consistent with the binding energy of O 2-in the TiO 2 lattice (Nasser, 2000;Kang and Lee, 2005).Other O 1 s spectra of metallic oxides, such as, VO 2 (530.0 eV), V 2 O 4 (530.0eV), V 2 O 5 (530.0 eV), and V 2 O 3 (530.3eV) (Moulder et al., 1995;NIST Chemistry WebBook, 2005)   oxides may be in the forms of VO 2 , V 2 O 4 , or V 2 O 5 but not in the form of V 2 O 3 .A shoulder to the main O 1s peak can be observed at high binding energy, which can be attributed to the hydroxyl groups or chemisorbed water molecules adsorbed on TiO 2 surface (Nagaveni et al., 2004).It is noted that the O 1s spectra of V/TiO 2 have a small binding energy shift compared with that of un-doped TiO 2 .The results can be ascribed to an increase in the ionic state of the oxygen bond causing the binding energies of all electronic states of oxygen to shift (Nasser, 2000).The V 2p 3/2 spectra of metallic V (512.4 eV), VO 2 (516.3 eV), V 2 O 3 (515.7 eV), V 2 O 4 (515.7 eV) and V 2 O 5 (517.4eV) are reported in the previous study (Moulder et al., 1995;NIST Chemistry WebBook, 2005).The photoelectron spectra of V 2p 3/2 can be decomposed into two peaks.The V 2p 3/2 main peak at 516.3 eV is assigned to be V (IV).A shoulder to the main V 2p 3/2 peak can be observed at 514.4 eV and is assigned to be V (III) (Moulder et al., 1995;NIST Chemistry WebBook, 2005).It has been reported that the formation of V 5+ is easy at high vanadium dopant concentration (Iketani et al., 2004).However, the V 2p 3/2 spectra of V 2 O 5 (517.4eV) are not observed from XPS data.TheV 4+ with radius of 0.63 Å can penetrate into the TiO 2 structure (Ti 4+ radius is 0.68 Å) for a coordination number of six.Consequently, the doped V ions may exist in the form of V (IV) and V-O-Ti within the TiO 2 lattices.The formed V-O-Ti would change the electron density due to its role as h + /e -traps resulting in differences in electro-negativity.The structure changes of TiO 2 upon V 4+ doping may partly explain enhanced photo-catalytic activity with V-doped photo-catalysts.

Photo-Catalytic Activities of V/TiO 2
The effects of vanadium doping concentration on 1, 2-DCE photo-degradation are shown in Fig. 5.When the UV lamp was turned on (t = 0), the production rate of photoinduced electrons and holes was suddenly excited, resulting in immediate 1, 2-DCE removal, although reaction rate tended to stabilize or decrease later on.The initial conversions for all systems appear the same, and then the conversions decay into a steady state within 20 minutes and remain for more than two hours.However, the conversion for V/Ti = 0.1 mol% is the highest in Fig. 5(a), but the conversion steadily declined until t = 50 min.The photo-catalytic activity of TiO 2 is the highest (η = 78%), followed by V/Ti = 0.1 mol% (η = 75%) > V/Ti = 0.5 mol% (η = 71%) > V/Ti = 1 mol% (η = 29%) at the end of experiments.It is indicated that the photo-catalytic activity of V/Ti = 1 mol% is lower.It is speculated that V 5+ ions have harmful effect on the photo-catalytic activity in the previous studies (Martin et al., 1994b;Iketani et al., 2004) and the formation of V 5+ ions is easy at high vanadium dopant concentration (Iketani et al., 2004).Therefore, our results consist with those of the reports.
To distinguish the relationship between photo-catalytic activity and vanadium contents further, the retention time and vanadium dopant concentration were adjusted.Referring to Fig. 5(b), the initial rate for all system appears the same and the conversion for 1, 2-DCE for photo-catalysts with V-doping systems all remained relatively stable.The photo-catalytic activity of V/TiO 2 with V/Ti = 0.01 mol% is the highest (η = 64%), followed by V/Ti = 0.001 mol% (η = 61%) > V/Ti = 0.1 mol% (η = 53%) > TiO 2 (η = 48%) at the end of experiments.The photo-catalytic activity is dependent on the vanadium content, and the optimum vanadium dopant concentration is 0.01 mol%.The result is consistent with that of a previous study (Iketani et al., 2004).
The photo-catalytic activity of photo-catalysts is determined by the rate of charge pair generation, charge trapping, charge release and migration, and charge recombination.Metal ion dopants can influence the photocatalytic activity of TiO 2 by acting as electron-hole traps and by retarding the electron-hole pair recombination rate.Somehow, the role of V must play an important role during the photo-catalytic reaction.It is speculated that the higher photo-catalytic activity of V/Ti = 0.01 mol%, is due to either one or combination of the following factors: (i) increased UV light absorption capability of V-TiO 2 compared to TiO 2 , or other V-TiO 2 ; (ii) form a lesser rutile phase (lower photo-catalytic activity) due to a small amount addition of V 4+ compared to other V-TiO 2 ; (iii) alleviate the surface poison phenomenon; and (iv) act as both h + /e -traps to reduce the recombination rate of h + /e -pairs during the photo-degradation.
V 4+ can act as both h + /e -traps to reduce the recombination rate of h + /e -pairs and enhance the photocatalytic activity, according to the following reactions (Choi et al., 1994;Klosek and Raftery, 2001): Charge-carrier generation (5) Charge release and migration The presence of molecular oxygen is necessary for the photochemical reaction, because O 2 can act to retard the h + /e -recombination rate.Further, O 2 -can also be involved in a variety of photochemical reactions.However, V (IV) in the TiO 2 lattice can also act as recombination centers for the h + /e -pairs if vanadium ion concentration is too high, according to the following reactions: Recombination The presence of vanadium species, especially at low levels of vanadium doping, retards the recombination rate of h + /e -pairs and enhance photo-catalytic activity.When vanadium doping concentration is large, the vanadium ions become recombination centers and reduce photo-catalytic activity.
Since the photo-catalytic reaction occurs on the surface of TiO 2 , surface area plays another important role in the photo-catalytic reaction.Referring to Table 2, although the addition of V (IV) decreases the surface area of V/TiO 2 , the photo-catalytic activity of V/TiO 2 (i.e., 0.01 mol%) is better than those of TiO 2 or V= 1mol%.Therefore, the surface area alone can not be used to explain the observed phenomena.Further, the SEM photographs and EDS results of reacted V/Ti = 1 mol% are shown in Fig. 6 and Table 3. Referring to Fig. 6 and Table 3, the adsorption of Cl element on the surface of V/TiO 2 is unobvious, so the poison phenomenon is unapparent on the surface of V/TiO 2 .The C element fraction is attributed to carbon deposits.Incomplete reacted 1, 2-DCE and the reaction intermediates including CH 2 Cl 2 and C 2 H 5 Cl may deposit on the surface of the photo-catalyst.Besides, it indicates that the actual V/Ti molecular ratio is equal to 0.98 mol% which is close to the added amount (1 mol%) in Table 3.Consequently, the role of V mentioned above may be responsible for the enhanced 1, 2-DCE photo-degradation, as compared to those from TiO 2 .

Intermediates of 1, 2-DCE Photo-Degradation Analysis
The intermediates and products of 1, 2-DCE photodegradation were analyzed by an on-line FTIR.The FTIR spectra of intermediates and products of 1, 2-DCE photodegradation are shown in Fig. 7. IR spectra were assigned with reference from the database (NIST Standard Reference Database, 2005).The CH 2 -Cl bending vibrations are observed in the range between 1,300 and 1,230 cm -1 and C-Cl stretching vibrations are observed in the region between 850 and 550 cm -1 .The C-Cl absorptions of intermediates produced in 1, 2-DCE e.g., CH 2 Cl 2 (739 cm -1 ), and C 2 H 5 Cl Wavenumber (cm -1 )  (670 cm -1 ) can be observed in the spectra.The characteristic peaks of intermediates overlap each other, so it is difficult to assign them accurately.Ou and Lo (2007) indicated the phosgene was formed when organic chloride was removal.CCl 2 O may be observed at 1,852 cm -1 by a FTIR.But the characteristic peaks of chlorinated products such as HCCl 3 (759 cm -1 ), CCl 4 (785 cm -1 ), and CCl 2 O (1,852 cm -1 ) were not observed in the IR spectra.The absorptions between 3,100 and 2,600 cm -1 are assigned to HCl.The products FTIR spectra of V/TiO 2 (0.01 and 0.001 mol%) are similar to those of nude TiO 2 in the characteristic peaks of HCl.However, the photo-oxidation of 1, 2-DCE by nude TiO 2 is apparently similar with V/TiO 2 .

CONCLUSIONS
The characterization of V/TiO 2 prepared by a sol-gel method was performed in this study.The photo-catalysts calcined at 500°C are spherical particles with a crystallite size about 10-20 nm and the crystal phase presents a mixture of anatase (dominant phase) and rutile.In addation, the anatase over rutile ratio of V/Ti = 0.01 mol% is the same as that of commercial Degussa P25 (80% anatase, 20% rutile).The UV absorbance with the V amount (0.01 mol%) exhibits the highest UV absorbance.However, since the V doping concentration is extremely low, its effect on visible light region is not clear in this study.From XPS data, the doped vanadium ions may exist in the forms of V 4+ or diffuse into the TiO 2 lattice to form the V-O-Ti bond.The formed V-O-Ti would change the electron density due to its role as h + /e -traps resulting in differences in electronegativity.The structure changes of TiO 2 upon V 4+ doping may partly explain enhanced photo-catalytic activity with V/TiO 2 .
The roles of V (IV) may alleviate the surface poison phenomenon and act as both h + /e -traps to reduce the recombination rate of h + /e -pairs.Referring to SEM images and EDS results, the poison phenomenon is unapparent on the surface of V/TiO 2 .Although the addition of vanadium ions decreases the surface area of V/TiO 2 , the photocatalytic activity of V/TiO 2 (i.e., 0.01 mol%) is better than those of synthesized un-doped TiO 2 .Consequently, the role of V may be responsible for the enhanced 1, 2-DCEphotodegradation, as compared to those from synthesized nude TiO 2 .

Table 1 .
Structure characteristics of various photo-catalysts.
are close to that of TiO 2 .The XPS results indicate that the vanadium

Table 2 .
Surface area of photo-catalysts with various vanadium ions doping concentrations.