Silver / Carbon Codoped Titanium Dioxide Photocatalyst for Improved Dye Degradation under Visible Light

Herein, we report the synthesis of quartz supported TiO2 photocatalysts codoped with carbon and silver through the hydrolysis of titanium tetrachloride followed by calcination at 500C.The prepared samples were characterized by UV-Vis diffuse reflectance spectroscopy, high resolution scanning electron microscopy (HRSEM), Raman spectroscopy, thermogravimetric analysis (TGA), X-ray photoelectron spectroscopy (XPS), and X-ray diffraction (XRD). Codoping of TiO2 with Ag and carbon resulted in an increase in the surface area of the photocatalyst and altered the ratio of the anatase to rutile phase. The absorption edge of all the doped TiO2 photocatalysts redshifted and the band gap was reduced. The lowest band gap of 1.95 eV was achieved by doping with 0.5% Ag. Doping TiO2 using carbon as the only dopant resulted in a quartz supported photocatalyst that showed greater photocatalytic activity towards methyl orange than undoped TiO2 and also all codoped TiO2 photocatalysts under visible light irradiation.


Introduction
The problem related to environmental pollution has triggered much research focused on the development of efficient methods for the removal of dangerous contaminants discharged into the environment.If these contaminants are not removed, they will have an adverse impact on human health and the environment.Photocatalysis using visible light responsive TiO 2 photocatalyst is an effective and economical way to reduce environmental pollution since it makes use of free sunlight and the contaminants can be mineralised completely into harmless products.The process involves the accumulation of pollutants on the photocatalyst surface which are then photodegraded by reactive species upon irradiation with light of appropriate wavelength.
Among the several semiconductors that are available such as CdS, ZnS, CdTe, ZnSe, CdSe, ZnO, WO 3 , CeO 2 , ZrO 2 , Nb 2 O 3 , Fe 2 O 3 , and SnO 2 , titanium dioxide is the most extensively used and studied photocatalyst due to its several advantages which include chemical stability, relative nontoxicity, and low cost [1].It has several applications which include (i) cleaning of oil spills [2,3], (ii) disinfection or antibacterial activity [4,5], (iii) usage as a white pigment, (iv) energy production [6], (v) water treatment [7,8], and (vi) air purification [7].The major drawback of the application of TiO 2 in photocatalysis is its wide band gap of 3.2 eV meaning that it can only absorb light of wavelength of less than 387 nm which is only 4% of the total incoming solar radiation.A lot of effort has been made to extend the photoresponse of the photocatalyst into the visible region which makes up 55% of the total solar radiation.
2 International Journal of Photoenergy However, the use of metal dopants has been found to result in localised d-levels in the TiO 2 band gap which can serve as recombination centres for charge carriers [22].It is important to note that, besides making the photocatalyst visible light responsive, there are other factors that exert a considerable influence on the performance of the photocatalyst and these include size, crystalline phase, pore volume, surface area, and exposed surface facets [23], but a lot of researchers have been focusing mainly on making use of visible light by reducing the band gap.The codoping of TiO 2 with a nonmetal and transition metal is therefore believed to be appropriate in narrowing the band gap with minimal recombination centres.The present paper reports the preparation of low band gap TiO 2 by codoping with carbon and silver using a sol-gel method.

Experimental
2.1.Materials and Chemicals.Special grade titanium (IV) tetrachloride 99% was purchased from Merck, Germany, silver nitrate (99.8%) was purchased from SAARCHEM, South Africa, and potassium hydroxide was purchased from Merck.All chemicals were analytical grade and used as received without further purification.Distilled water was used in all preparations of solutions.

Catalyst Preparation.
A modified version of the sol-gel procedure reported in literature [20] was used to prepare the silver/carbon codoped TiO 2 photocatalysts with different contents of Ag.In a typical experiment, titanium tetrachloride (6.0 mL) was added dropwise to a 100 mL of distilled water in a flask immersed in an ice bath with vigorous stirring.An appropriate amount of glucose was added to get a final concentration of 0.01 M of glucose in the solution.This procedure was repeated to prepare six different sets of solutions.Suitable amounts of silver solution were added to each of the six solution sets so as to get 0.2, 0.5, 0.7, and 1.0 mol percent of Ag, respectively, in the solutions.The solutions were then heated for 30 minutes at 95 ∘ C and then allowed to cool to room temperature.After cooling, the pH was adjusted to 8.0 using potassium hydroxide and then heated again at 95 ∘ C for another 30 minutes.The precipitate was separated from the solution by centrifugation, washed with deionised water to remove excess chloride ions, dried at 90 ∘ C, and then finally calcined at 500 ∘ C in air for 2 hours to give Ag/C codoped TiO 2 .The undoped TiO 2 photocatalyst was also prepared using the same procedure but without the addition of glucose and silver.Ag doped TiO 2 photocatalyst was prepared using the same procedure.

Deposition of TiO 2 Photocatalysts on
Quartz.The prepared C-doped and Ag/C codoped TiO 2 photocatalysts were deposited on quartz supports using a modified version of the methodology reported by Markelonis and coworkers [24].The method has the advantages of being fast and simple compared to other currently known nanoparticle deposition techniques.A desired amount of the prepared TiO 2 (0.02 g) photocatalyst was mixed with absolute ethanol (15 mL) and then sonicated for five minutes to make sure that the TiO 2 nanoparticles were well dispersed.The etched and 3-mercaptopropyl trimethoxysilane (MPTMS) treated quartz substrate to be coated was placed in a centrifuge vial with one side covered with sellotape so that only one side could be coated.The sonicated TiO 2 nanoparticles were then poured into the centrifuge vials and centrifuged at 3500 rpm until a significant amount was deposited.The desired amount was determined by weighing the quartz support before and after deposition.After centrifugation, the supernatant was pipetted out and the quartz substrate with the deposited TiO 2 nanoparticles was then removed and allowed to dry in the air.

Characterization.
The diffuse reflectance spectra (DRS) of the samples were acquired from 800 nm to 200 nm at room temperature using a diffuse reflectance attachment of a Cary 500 UV-Vis-NIR spectrophotometer.Fourier transform infrared (FTIR) spectra were obtained on a Perkin-Elmer (FTIR Spectrum 200) spectrophotometer.The powdered samples were mixed with KBr, pressed into a pellet, and then analysed.The crystallographic characteristics of the prepared samples were analysed by a powder X-ray diffractometry (XRD) technique using a Bruker D8 Advance instrument with a Cu-K1 ( -1.54060) X-ray Tube and a LynxEye detector.The Raman spectra were obtained on a Czerny-Turner micro-Raman spectrometer equipped with a cooled charged coupled device.The detector and beam path were set at −50 ∘ C and 50x, respectively.A red laser diode with a wavelength of 785 nm was used for the excitation of the samples.The specific surface areas of the powdered Ag/C codoped TiO 2 photocatalysts were obtained by N 2 adsorption before deposition on quartz using a Micromeritics ASAP 2020 surface area and porosity analyser.X-ray photoelectron spectroscopy (XPS) was done using a Kratos Axis Ultra DLD, using an Al (monochromatic) anode, equipped with charge neutralizer and the operating pressure was kept below 5 × 10 −9 torr.

Photocatalytic Activity Evaluation.
The photocatalytic activity of the synthesised Ag/C codoped TiO 2 photocatalysts was evaluated under visible light irradiation using methyl orange as a model pollutant.All photodegradation experiments were conducted at a temperature of 25 ∘ C and at pH 7. In a typical experiment, the 2.5 cm × 3.5 cm photocatalyst coated quartz slide was immersed in 100 mL of 10 ppm methyl orange solution in a flask in such a way that there was room at the bottom of the reactor by means of a perforated plastic stand for the magnetic stirrer to rotate without scratching the photocatalyst film.The flask was wrapped with aluminium foil and then placed in the dark for an hour for adsorption and desorption equilibrium to take place.The system was then irradiated with visible light from an Oriel Newport solar simulator fitted with a 500 W xenon lamp.The solar simulator was Model 91150V supplied by Oriel Instruments, USA.To detect the changes in methyl orange concentration with time, aliquots of methyl orange were taken after every 30 minutes and analysed by a UV-Vis spectrometer at a wavelength of 468 nm.

FTIR Analysis.
The FTIR analysis was done to study the changes in the surface of the codoped photocatalysts as the Ag dopant level was increased and also to determine the functional groups.The FTIR spectra in the region between 400 cm −1 and 4000 cm −1 of undoped TiO 2 , C-doped TiO 2 , and Ag/C codoped TiO 2 photocatalysts are shown in Figure 1 and they all have similar peaks but with different intensities.The peaks around 3450 cm −1 and 1655 cm −1 were assigned to the stretch vibration and characteristic frequency of O-H groups bending of adsorbed water, respectively [25].These were more intense in the spectrum of the photocatalyst with Ag loading of 0.5%.This probably implies that this particular photocatalyst adsorbed water more than the other photocatalysts which could be useful in photocatalytic reactions because the adsorbed water can be attacked by positive holes to produce hydroxyl radicals.The pronounced broad band in the range 400-900 cm −1 was assigned to the stretching vibration of the Ti-O bond and it was common in the spectra of all samples.The peak at 2358 cm −1 due to the presence of carbon species [26] is found in the spectra of all samples except TiO 2 that had Ag content of 0.0% and also undoped TiO 2 .The undoped TiO 2 is the only photocatalyst that had a peak at 1557 cm −1 .
When the XRD pattern of undoped TiO 2 photocatalyst is compared with the XRD pattern of doped TiO 2 photocatalysts, the relative intensities of some peaks and peak areas, for instance, at 2 = 47.8 ∘ , were altered due to lattice distortions and changes in the degree of crystallinity caused by the conditions applied.
The XRD results showed that Ag and carbon codoping of TiO 2 altered the ratio of anatase to rutile phase present.For instance, in the XRD pattern of undoped TiO 2 , the most intense peak at 2 = 25.1 ∘ due to the anatase phase reduced significantly in the XRD pattern of Ag/C codoped TiO 2 photocatalysts and it completely disappeared in the XRD pattern of C-doped TiO 2 photocatalyst.The peaks at 2 = 27.3 ∘ , 36.2 ∘ , and 41.1 ∘ that are assigned to the rutile phase were of low intensities in the XRD pattern of undoped TiO 2 but increased in intensity in the XRD patterns of all doped TiO 2 photocatalysts indicating partial phase transformation had occurred by doping with Ag or carbon.There was a new peak at 2 value of 43.8 ∘ due to rutile which was nonexistent in the XRD pattern of undoped TiO 2 but present in the XRD pattern of all doped TiO 2 samples.The major phase in the C-doped TiO 2 (0.0% Ag) was rutile as shown in its XRD pattern in Figure 2 with very intense peaks due to rutile when compared to other XRD patterns.Two new peaks at 2 = 43.6 ∘ and 63.6 ∘ assigned to the rutile phase and anatase phase, respectively, appeared in the XRD patterns of all doped samples but were not visible in the XRD patterns of undoped TiO 2 .
The crystallographic parameters of the synthesised undoped TiO 2 obtained from the XRD instrument were as follows: crystal system: tetragonal, space group: P42/mm, space group number: 136, and lattice constants of ( Å), ( Å), ).From this information it shows that the prepared TiO 2 had no brookite phase but only anatase and rutile.
To estimate the anatase percentage (%), the following equation was used for calculations [28].
where   is the peak intensity of rutile at 2 theta' value of 27.4 ∘ .  is the peak intensity of anatase at 2 theta' value of 25.3 ∘ .The calculated percentages of anatase and rutile phases in undoped TiO 2 and Ag/C codoped TiO 2 are shown in Table 1.When compared to undoped TiO 2 , the percentage of anatase decreased in all doped samples due to phase transformation to rutile which was related to Ag or C doping.
The highest percentage of the anatase phase of 88.13% and the lowest percentage of rutile of 11.88% were found in the undoped TiO 2 photocatalyst which was also calcined at 500 ∘ C. All the Ag/C codoped TiO 2 photocatalysts had higher percentages of the rutile phase than the anatase phase.The highest anatase to rutile phase transformation occurred at 500 ∘ C when TiO 2 was doped with carbon only, with less rutile formed in Ag/C codoped TiO 2 (0.7% Ag) photocatalyst with 2.20% anatase and 97.80% rutile.
Well defined peaks were observed at 145 cm −1 (E g ), 244 cm −1 , 404 cm −1 (B 1g ) prominent in undoped TiO 2 , 453 cm −1 (E g ), 525 cm −1 (B 1g ), 617 cm −1 (A 1g ), and 641 cm −1 (E g ) as shown in Figure 3.It is clear that the strongest band for the anatase is at 145 cm −1 in the spectra of undoped TiO 2 which slightly shifted to 147 cm −1 in the spectra of doped TiO 2 .The presence of the peak at 244 cm −1 due to the rutile phase in all doped TiO 2 samples except the undoped TiO 2 shows that the high temperature of calcination or  introduction of impurities in the structure of TiO 2 could have caused the anatase to rutile phase transformation which is in agreement with XRD results.This is also supported by the big reduction in intensities of the anatase peaks at 404 cm −1 and 644 cm −1 and a huge increase in the intensities of the rutile peaks at 453 cm −1 and 617 cm −1 in the spectra of doped TiO 2 .The vibrational mode of the rutile phase at 143 cm −1 could not be detected probably due to the superimposition with the anatase phase at 144 cm −1 .When carbon was used as the only dopant, the E g peak at 144 cm −1 nearly disappeared and all its major peaks were due to the rutile phase.This peak is very sensitive to the vibrational mode of the O-Ti-O bond [31].4. When compared to undoped TiO 2 photocatalyst, the absorption edge of all the Ag/C codoped photocatalysts redshifted and the highest shift was observed for the photocatalyst with 0.5% Ag loading.When the Ag content was increased from 0.5% to 1.0%, the absorbance blueshifted implying that the optimum loading for Ag in this study to obtain a catalyst that could absorb the most visible light was 0.5%.

UV-Vis Diffuse
Reflectance.The diffuse reflectance spectra show a very clear and significant shift of the absorption edges of the doped photocatalysts to higher wavelengths when compared to the undoped TiO 2 photocatalyst as shown Figure 5.The reason for the redshift allowing visible light absorption by the photocatalysts could be the introduction   of impurity levels with Ag doping [32].The electron transfer from one of the additional impurity energy levels within the band gap of the semiconductor requires low photon energy.There were very small differences between the absorption edges of the photocatalysts with 0.2%, 0.7%, and 1.0% Ag loading.A significant shift was observed for the photocatalyst  with Ag loading of 0.5%.Of all the doped photocatalysts, the one doped with carbon as the only dopant (0.0% Ag) showed the least shift, which shows that codoping is better than the use of one dopant in enhancing visible light absorption.

Band Gap Determination.
The band gaps were determined using Tauc plots which are plots of [() * ℎV] 1/2 versus photon energy.The extrapolation to the horizontal axis (energy axis) gave an estimate of the value of the band gap (E g ).The Tauc, Davis, and Mott expression is shown in where  is the absorption coefficient, ℎ is Plank's constant, E g is the band gap,  is the proportionality constant, V is the frequency of vibration, and  is a constant which denotes the nature of the sample preparation.In this study the value of  used for the calculations was 2 since the absorption of TiO 2 has an indirect transition.The Tauc plots of Ag/C codoped TiO 2 photocatalysts are shown in Figure 6.Undoped TiO 2 photocatalyst gave the most well defined absorption edge that made it quite easy to make an extrapolation and determine the band gap more accurately (ca.3.06 eV).The band gaps of all the Ag/C codoped TiO 2 photocatalysts were between 1.95 eV and 2.90 eV thus lower than that of the undoped TiO 2 .This band gap narrowing could be attributed to the substitution of carbon and/or metal Ag for oxygen.
There was a decrease in the band gap from 2.83 eV to 1.95 eV as the Ag loading was increased from 0.2% to 0.5% and any further increase in the Ag loading caused band gap widening (Table 2).The possible reason for the increase in band gap with an increase in Ag content was the dominance of the d-d transitions over the sp-d transitions.The optimum loading of silver in this study for codoping TiO 2 with silver and carbon was found to be 0.5%.It is important to note that new occupied bands could have been formed as a result of the hybridization of Ag 4d and Ti 3d states.

XPS Analysis.
The presence of carbon and silver in TiO 2 was revealed by XPS analysis.The high resolution XPS scans over the Ti 2p, O 1s, and C 1s in Ag/C codoped TiO 2 (0.5% Ag) photocatalyst are shown in Figure 7.
The spectra show that the Ag/C codoped TiO 2 photocatalyst contained the elements carbon, titanium, oxygen, and silver.The XPS spectrum of Ti 2p (Figure 7(b)) shows two peaks at binding energies of 457.5 eV and 463.5 eV which were attributed to the spin orbit coupling of Ti 2p 3/2 and Ti 2p 1/2 , respectively, and the binding energy separation between these peaks was 6 eV which is characteristic for TiO 2 [33].The peak at 564 eV was due to Ti 2s (Figure 7(a)) and was not as intense as the Ti 2p peak.The peak intensity of Ti 4+ was much higher than that of other species implying that, in the TiO 2 lattice, Ti 4+ was dominant.When Ag + replaces Ti in the lattice, it causes oxygen vacancies which result in some changes of Ti 4+ ions into Ti 3+ ions [34].From the survey scan, only 0.13% of Ag was detected in the codoped TiO 2 and this can be The O (1s) XPS spectrum (Figure 7(c)) consists of one major peak at 528 eV and a shoulder at 530.5 eV which corresponds to oxygen bound to TiO 2 (O-Ti-O) and surface hydroxyl groups (T-OH), respectively.The C 1s of the prepared Ag/C codoped (0.5% Ag) photocatalyst revealed two peaks at binding energies of 283.2 eV and 287.3 eV (Figure 7(d)).The peak at 283.2 eV can be assigned to carbon that substituted an oxygen atom in the crystal lattice of TiO 2 forming C-Ti-O bond [35].In this study the peak at 287.3 eV can be attributed to C-O bond; hence the XPS spectrum proved that some C was doped into the TiO 2 lattice.
3.6.BET Surface Analysis.The differences in the surface area and porosity of particles of the same material, which may otherwise have the same physical characteristics, can greatly influence its performance characteristics.The highest surface area of 83.13 m 2 g −1 was achieved when carbon was used as the only dopant implying that carbon doping increases the surface area of TiO 2 (Table 3).The surface area of the Ag/C codoped TiO 2 photocatalysts increased slightly with an increase in the silver content and decreased with silver contents above 0.7%.It is believed that a photocatalyst with high surface area is desirable in photocatalysis since high surface areas allow greater adsorption of pollutants resulting in enhancement of photocatalytic activity.All the samples started showing a mass loss at around 100 ∘ C due to loss of adsorbed water on the TiO 2 surface.The highest weight loss of 3.83% was seen in undoped TiO 2 .There was a gradual increase in total weight loss at 800 ∘ C with an increase in Ag content.The TiO 2 with 0.0% Ag showed a total weight loss of 1.46% whereas the TiO 2 photocatalyst with 1.0% Ag showed 2.36% weight loss.Since the total mass loss in all cases was less than 3.83%, it suggests that the photocatalysts were thermally stable and most of the carbonaceous material from glucose (C-dopant source) was lost during the pyrolysis process, since the decomposition process of glucose started at 120 ∘ C and pyrolysis started at 320 ∘ C up to 800 ∘ C.

Photocatalytic Activity Evaluation
The highest photodegradation of methyl orange under visible light was achieved by C-doped TiO 2 (0.0% Ag) photocatalyst probably due to the high surface area for adsorption of the contaminant and also enhanced visible light absorption.There was an increase in the photodegradation of methyl orange from 47.61% to 52.07% when the Ag content was increased from 0.2% to 0.5% and it decreased with a further increase in Ag content of 0.7% (Figure 9).The optimum loading of Ag in the codoped TiO 2 for efficient photocatalysis in this study was found to be 0.5%.The reason for the high photodegradation rates from this photocatalyst could be the low band gap of 1.95 eV allowing efficient visible light absorption.It is important to note that the material with a lower band gap will not always have higher photocatalytic activity as expected.For instance, the photocatalytic activity of Ag/C-TiO 2 (0.7%) with a calculated band gap of 2.53 eV International Journal of Photoenergy  was lower than that of Ag/C-TiO 2 (1.0%) with a band gap of 2.72 eV which was unexpected on the basis of band gaps.This observation points out that band gap narrowing could improve visible light absorption but without improving photocatalytic activity under visible light irradiation.This implies that there could be other factors which override band gap narrowing that affect photocatalytic activity of TiO 2 including electron-hole recombination rates, hydrophilicity, and enhanced adsorption due to different surface areas and surface properties.
The general high photocatalytic activity of all the doped TiO 2 photocatalysts could be explained by the formation of oxygen vacancies induced by Ag + ions and trapping of electrons by Ag + through the introduction of allowed electronic states below the TiO 2 conduction band [36].This resulted in enhanced visible light absorption and reduced recombination rates of the charge carriers.

Conclusions
Ag/C codoped TiO 2 photocatalysts were successfully prepared by a simple sol-gel technique and deposited on quartz substrates.The study showed that a photocatalyst with a low band gap does not always have better photocatalytic activity than wide band gap photocatalysts under visible light irradiation as expected.TiO 2 photocatalyst with a significant low band gap of about 1.95 eV can be prepared by codoping with carbon and silver (0.5% Ag) and codoping does not always give the best photocatalyst as expected when compared to doping with a single element.The study also showed that there are other factors that override band gap narrowing in enhancing the overall efficiency of the photocatalyst.

Figure 2 :
Figure 2: XRD patterns of undoped TiO 2 and Ag/C codoped TiO 2 photocatalyst powders with different contents of Ag prepared by a solo-gel method.

3. 7 .
Thermal Properties.The powdered samples (10 mg) were heated from 20 ∘ C to 800 ∘ C at a heating rate of 20 ∘ C/min in

Figure 9 :
Figure 9: Photodegradation of 10 ppm methyl orange solution using Ag/C codoped TiO 2 photocatalysts with different contents of Ag under visible light irradiation at 25 ∘ C and pH 7.

Table 2 :
Band gaps of undoped TiO 2 , Ag doped TiO 2 , and Ag/C codoped TiO 2 with different contents of Ag.

Table 3 :
BET surface area, pore volume, and pore size of undoped TiO 2 and Ag/C codoped TiO 2 photocatalysts.by the radius of Ag + (ca.126 Å) which is larger than that of Ti 4+ (ca.68 Å) thus making replacement of Ag into the TiO 2 lattice difficult and requiring energy. explained