Visible Light Excited Catalysis and Reusability Performances of TiO 2 @ Pr : Y 2 SiO 5 Upconversion Materials

1 Jiangsu Engineering Technology Research Center of Environmental Cleaning Materials (ECM), School of Environmental Sciences and Engineering, Nanjing University of Information Science and Technology, Nanjing, China 2Jiangsu Key Laboratory of Chemical Pollution Control and Resources Reuse, School of Environmental and Biological Engineering, Nanjing University of Science and Technology, Nanjing 210094, China


Introduction
Photocatalysts responding to visible light have been studied in environment for many years, and lots of interesting research results are booming out on the photodegradation of pollutants.As one of the best photocatalyst, TiO 2 has been studied extensively on the decomposition of inorganic and organic pollutants and toxic materials [1][2][3], as well as the application for water purification [4,5].Since the energy gap of anatase TiO 2 is about 3.2 eV, only high energy UV light with wavelength less than 387 nm can excite the electron transition of TiO 2 [6].However, only about 4% in the solar light of UV light can be absorbed by nanometer anatase TiO 2 .Many studies focus on decreasing energy gap of anatase TiO 2 , including surface modification [7,8], doping with other metal ions, nonmetal ions and semiconductors [9][10][11][12][13], and oxygen vacancy generating [14], which have been researched to develop TiO 2 as a visible light responding photocatalyst.
Upconversion material, such as Yttrium Silicates (Y 2 SiO 5 ), is a kind of material which can convert lower frequency light to upper frequency light.This unique property enables upconversion material with a great potential application in many fields.To get high luminescence intensity, activation methods were widely researched for Y 2 SiO 5 luminescent material, including doping with other metal ions or activating agents [15,16].Y 2 SiO 5 has a particular geometry which leads to a possible replacement of Y by other element ions, especially by rare earth ions [17].Praseodymium ion Pr(III) was doped into Y 2 SiO 5 and high luminescence intensity was gotten, because praseodymium ion has more suitable energy levels and longer excited state lifetimes than Y [18].This means that, under the irradiation of visible light, UV light could be excited out by upconversion materials, which can be used by nanometer TiO 2 .Therefore, the combination of nanometer anatase TiO 2 with upconversion materials could hopefully make more efficient use of solar energy in practical applications and provide a wide use of TiO 2 in photodegradation fields.Furthermore, with the combination of TiO 2 , lithium ion (Li + ) has been used to dope Pr:Y 2 SiO 5 upconversion materials and high photocatalysis performances on nitrobenzene wastewater were gotten under visible light [18].
However, it was found that upconversion materials were seldom used as a photocatalyst independently for wastewater treatment.In order to improve the photocatalysis ability of TiO 2 excited by visible light or solar, based on the upconversion materials, various forms of TiO 2 and support materials have been reported in recent years.Hydrogen is produced efficiently from aqueous methanol solution by using Pt supported TiO 2 combined with upconversion luminescence agent (Er 3+ :Y 3 Al 5 O 12 ) as the visible light photocatalysts [19,20].Er 3+ doped TiO 2 thin films can present better response to infrared light upconversion [21].With the assist of upconversion luminescent materials, TiO 2 solar photocatalysts were applied to degrade contaminants to purify water [22].For the degradation of organic dyes, TiO 2 coating upconversion luminescence agent [23] and rutile TiO 2 catalyst doped with upconversion luminescence agent [24] both showed excellent performances under visible light irradiation.
In the treatment process, photocatalysts powders were usually dispersed in wastewater directly [18].Although this process enables full contact and high degradation efficiency of the catalysts particles on the pollutants, the recovery of the catalyst powder becomes a new problem [18,25].In fact, incomplete recovery of the catalyst particles would lead to secondary pollution for water treatment.Nevertheless, it is very difficult to recover those catalyst particles totally from the treated water, especially for nanometer particles.Reasonably, catalysts carrier in bulk size were used to load catalyst particles and made reusing of catalysts possible [26,27].
The doping of Li + was very limited to enhance the photocatalysis for Pr:Y 2 SiO 5 upconversion materials; the degradation rate was only about 80% after 6 h of treatment.On the other hand, at the same treatment time, the upconversion material without Li + had gotten a degradation rate of 78%, as shown in Figure 6 in [18].In this study, Pr(III) doped Y 2 SiO 5 upconversion materials and anatase TiO 2 nanofilm coated Pr:Y 2 SiO 5 composite have been prepared to get high photodegradation efficiency on pollutants.Furthermore, glass fiber filters, which are inert, inexpensive, and easily available, have been applied to load the upconversion composite materials by using a simple dip-coating method.The degradation performances and reusability of obtained catalysts under visible light have been tested by using nitrobenzene as typical waste in water.
The composite catalysts of TiO 2 @Pr:Y 2 SiO 5 , named as TPY, were prepared by adding the as-prepared Pr:Y 2 SiO 5 powder, which was ground and ultrasonic-dispersed in absolute alcohol for 30 min, to the mixture of titanium tetrabutoxide, absolute alcohol, and distilled water with a volume ratio of 1 : 7 : 2. The pH of mixture was adjusted to 2.5 by using nitric acid.After stirring for 1 h, the sol suspended Pr:Y 2 SiO 5 particles were put into a water bath at 70 ∘ C to form a white gel.Then, the gel was dried in an oven at 80 ∘ C and milled into powder.At last, the powder was calcined at a temperature of 500 ∘ C for 2 h in a muffle furnace to get the product TiO 2 @Pr:Y 2 SiO 5 (TPY) composite catalyst.The TiO 2 ratios, 1%, 2%, 3%, 4%, and 5%, were adjusted by changing the added volume of titanium tetrabutoxide.We note here that when the sample name TiO 2 @Pr:Y 2 SiO 5 (TPY) is mentioned in this paper, it mainly means 4% TiO 2 in the materials except for the comparing study on the content of TiO 2 .Titanium dioxide powder was prepared by following the same procedure preparing TiO 2 @Pr:Y 2 SiO 5 except for adding Pr:Y 2 SiO 5 particles in the sol.
The glass fiber filter (GFF) loaded TiO 2 @Pr:Y 2 SiO 5 or TiO 2 , named as TPY@GFF and TiO 2 @GFF, respectively, was prepared by using a dip-coating method through dipping GFF into the sol suspended Pr:Y 2 SiO 5 particles or the sol of TiO 2 precursor.Then the coated GFFs were heat-treated by drying at 80 ∘ C and calcining at 500 ∘ C sequentially.Before the dip-coating process, the glass fiber filters were washed by HNO 3 (1 : 1, vol) and diluted water for three times in turns to eliminate impurities in GFF and make the coating firmly.

Property Testing of Samples. X-ray diffractometer (D8
Advance, Bruker Corporation, German) was used to characterize crystal form and grain size of the samples.A scanning electron microscopy (Hitachi S4800, SEM, Japan) was employed to characterize the microimages and particle size of the samples.The BET specific surface areas of samples were tested after evacuating for 6 h at 160 ∘ C by using a specific surface area and pore size analyzer (V-Sorb2800P, Gold APP Instruments Corporation, China).The upconversion luminescence of the materials was tested by using a fluorescence spectrometer (FL3-TCSPC, Horiba Jobin Yvon Corporation, France).The exciting parameters were selected as 488 nm of the excitation wavelength, 370 nm of the optical filters, and 1 nm of the slit [29,30].All the samples are cut into small pieces or milled into powder to fit the sample cell of the fluorescence spectrometer.
Nitrobenzene wastewater, 5 mg/L, was used as a target pollutant to test the photodegradation performances of the as-prepared upconversion composite nanomaterials.After dipping the GFF loaded with catalysts into the wastewater or dispersing the photocatalysts powder into the wastewater, half an hour of dark environment was experienced to reach an adsorbing balance of the pollutants on the catalysts materials.Then, a triphosphor tube light 100 W with an ultraviolet filter was turned on as a visible light source to excite the upconversion nanomaterials and photocatalysts.The degradation rate of nitrobenzene was calculated by comparing the ultraviolet absorption values at the wavelength of 267 nm to the original values of the nitrobenzene solution.The ultraviolet absorption values were tested by using an ultraviolet-visible spectrophotometer.When treated for a certain time, about 3 ml of the solution was took out from the reaction vessel and injected into the cuvette of the ultravioletvisible spectrophotometer and put back to the vessel after test.The relationship of the nitrobenzene concentrations (), in the range of 0.5-10 mg/L, with the ultraviolet absorption value (), was determined by a linear equation  = 9.06 − 0.312, with a correlation  2 of 0.99982.

Physical Properties
Testing for the Samples.The crystal structure and crystal size play important roles in the photocatalysis properties for TiO 2 particles, as well as the light conversion performances for upconversion nanomaterials [28,31].Figure 1 shows the XRD patterns of the as-prepared samples TiO 2 @Pr:Y 2 SiO 5 , Pr:Y 2 SiO 5 , and TiO 2 .The patterns of Pr:Y 2 SiO 5 and TiO 2 present good accordance with the standard PDF card of #41-0004 and #21-1272, respectively.These also insist that sample Pr:Y 2 SiO 5 is of low temperature phase X1 molecular configuration and TiO 2 belongs to anatase phase, of which the two kinds of crystal structures are both suitable for light response.Meanwhile, TiO 2 @Pr:Y 2 SiO 5 shows all the characteristic patterns of anatase phase TiO 2 and along with a peak at about 2 theta of 31 ∘ in which most of characteristic patterns of Pr:Y 2 SiO 5 are located.It indicates that TiO 2 and Pr:Y 2 SiO 5 are two independent materials coexisting in the composite of TiO 2 @Pr:Y 2 SiO 5 .Titanium dioxide probably in film form coats the surface of Pr:Y 2 SiO 5 particles, which also can be suggested by the SEM photos of the samples, as shown in Figure 2.
According to the XRD patterns in Figure 1, the crystal size of Pr:Y 2 SiO 5 is 41.7 nm at 2 angle 30.83 ∘ , which are calculated by the Scherrer equation [17]:  = /( cos ), where  is the crystal size (nm);  is 0.89, the Scherrer constant;  is the diffraction angle at which the diffraction peaks are located ( ∘ );  is the full width at half maximum (FWHM) of the main diffraction peaks (rad), which is located at the 2 angle of 30.83 ∘ in this situation;  is 0.154056 nm, the X-ray wavelength.Meanwhile, the crystal sizes of TiO 2 and TiO 2 in the TiO 2 @Pr:Y 2 SiO 5 composite sample are calculated, 33.1 nm and 22.5 nm, respectively, at 2 angle 25.1 ∘ .
However, as shown in Figure 2, SEM particle sizes are much bigger than those crystal sizes calculated by XRD.It indicates that every single particle is composed of several crystals which makes it difficult to disperse the Pr:Y 2 SiO 5 powders fully in the preparation process.The grains are about 0.5 to 1.0 m in both of the two images of Figure 2. The particle sizes in image (b) are a little bigger than that in image (a), which insists on the coating of TiO 2 on the surface of Pr:Y 2 SiO 5 grains.Nevertheless, energy dispersive X-ray spectrometer (EDS) testing on the sample surface of Figure 2(b) shows that Ti and O are the majority elements, while Si and Y are the minority elements.It also suggests that TiO 2 film gives intact coating on Pr:Y 2 SiO 5 particles.Furthermore, it is difficult to find element information of Pr in the EDS testing, which probably has gotten into the crystal lattice and replace the host atoms of Y 2 SiO 5 [17].
Figure 3 shows SEM photos of glass fiber filters (GFF) before and after loading the photocatalysts.In the microimage of TiO 2 @GFF, small TiO 2 particles are well loaded in the interspace of GFF, as shown in Figure 3(b).However, although many small particles are also loaded in the interspace of GFF, there are some big particles present on the GFF surface, as shown in Figures 3(c) and 3(d).It is no doubt that these big particles would drop out firstly during the treatment process, because the adhesive forces of big particles on the surface are less than that of small particles in the interspace.Despite that, there are still many big particles loading on the surface firmly even after having been reused for 6 times, as shown in Figure 3(d).It suggested that the simple combination of sol-gel method and dip-coating method could be useful for loading photocatalysts precursor on carrier.
Generally, the specific surface area is a key parameter on the adsorption properties of porous materials.As a result, it highly affects the degradation performances of catalysts since the pollutants must be adsorbed onto or be getting very close to the catalysts before the start of the degradation process.The adsorption isotherm linear graphs of the as-prepared catalysts and the GFF before and after loading TPY catalysts are shown in Figure 4.All the three curves in Figure 4(a) show the characteristics of type I adsorption isotherm, which suggests that the adsorption performances of the catalysts, TiO 2 , Pr:Y 2 SiO 5 , and TiO 2 @Pr:Y 2 SiO 5 , are mainly attributed to micropores.The specific surface areas of the three catalysts are 231.58m 2 /g, 12.34 m 2 /g, and 127.85 m 2 /g, respectively, which show good accordance with the curves in Figure 4(a).On the other hand, the specific surface area of GFF is only 4.61 m 2 /g, and its adsorption isotherm line shows adsorption isotherm characteristics of type V. Nevertheless, after loading TPY catalyst, the adsorption line of TPY@GFF is changed to type II adsorption isotherm.It is interesting that this line is very similar to the curve of TPY in the relative pressure range of 0-0.15, while presenting similar characteristics of GFF curve in the relative pressure range of 0.15-0.30.This suggests that the adsorption performances of TPY@GFF are composed of two stages: micropore adsorption of TPY catalysts dominated at the stages of low relative pressure and macropore adsorption of GFF is mainly at the stage of high relative pressure.The same features also can be found on the adsorption isotherm line of TPY@GFF used 6 times, as shown in Figure 4(b).It is easy to be understood that, because of the dropping out of the catalysts particles in the treatment procedures, the specific surface area of TPY@GFF used 6 times would be a little less than that unused, which are 12.58 m 2 /g and 13.07 m 2 /g, respectively.This presents good evidence on the excellent recycle usability performances of the TPY@GFF, which also has been suggested by the SEM images in Figure 3.

Upconversion Luminescence of Samples.
Figure 5 shows the upconversion emission spectra of Pr:Y 2 SiO 5 with different concentrations of TiO 2 , which are excited by light with the wavelength of 488 nm.Each of the samples shows an obvious upconversion emission peak in the wavelength range of 290 nm to 340 nm with a peak value at 312 nm.It can be found that the upconversion luminescence intensities are decreasing with the coating ratios of TiO 2 , 0%, 1%, 2%, 3%, 4%, and 5%.This should be attributed to the semiconductor characteristics of TiO 2 .It is known that titanium dioxide is a kind of -type semiconductor material with a band gap of 3.2 eV (for anatase).When it is exposed to ultraviolet light whose wavelength is less than 387.5 nm, energy would be added to valence electron and it would be excited by photon and the electrons would leap from the valence band (VB) to the conduction band (CB), where it can move freely around the crystal in the form of photoelectron [32].As a result, a hole is left behind in the valence band.On the occasion of nanoscale TiO 2 coating on the upconversion material Pr:Y 2 SiO 5 , the luminescence emitted by Pr:Y 2 SiO 5 is absorbed in situ by the TiO 2 coating film which led to a sharp decreasing of the luminescence intensities.The more of TiO 2 film is coated, the less of the luminescence intensity can be tested.When the coating amount of TiO 2 film is up to 4%, the intensity of the emission spectra is very low, as shown in Figure 5.The luminescence peaks almost cannot be found on the emission spectra of the sample with 5% of TiO 2 .It means that too thick TiO 2 coating film would not only block the transmission of light onto the upconversion material but also absorb out the luminescence that the upconversion material Pr:Y 2 SiO 5 emitted.Therefore, the coating amount of 4% could be a balance point or an optimum value for the composite of TiO 2 @Pr:Y 2 SiO 5 nanomaterials.
After being loaded on the GFF, the catalyst TiO 2 @Pr:Y 2 SiO 5 , with 4% of TiO 2 , shows notable decreasing on the upconversion emission intensity, as shown in Figure 6.This could be attributed to the penetrating of the catalyst particles into the interspace of glass fiber filter during the dip-coating process, which decreases the emission intensity what can be detected.Besides, the upconversion emission spectra of TYP@GFF before and after being used 6 times have also been tested, as shown in Figure 6, to check the recycle usability performances of the sample.After being used 6 times, little decreasing of the upconversion emission intensity of the TYP@GFF materials can be seen.It could be verified that TYP@GFF would probably have excellent reusing performances, which also have been suggested by the SEM photos in Figure 3 and the adsorption isotherms in Figure 4.

Visible Light Photodegradation on Nitrobenzene Wastewater.
Coating on the surface would hopefully improve catalysis performances of base particles because of the synergistic effects of the coating materials and base materials, although the process would not change the crystal structure or crystal size of base particles.The synergistic effects improvement of TiO 2 with Pr:Y 2 SiO 5 has been tested, as shown in Figure 7.For these five samples, the photodegradation performances show high consistency with the intensities of upconversion luminescence which they emitted, as shown in Figure 5.Meanwhile, the photodegradation rates increase gradually with the TiO 2 coating rate increase from 0% to 4%.It can be deduced that too less coating of TiO 2 film on the Pr:Y 2 SiO 5 particles would lead to inadequate use of violet light that the upconversion materials converted.For the sample of Pr:Y 2 SiO 5 (0% of TiO 2 ), it shows much more poor photodegradation performances than those samples coated with TiO 2 .It is because of the absence of TiO 2 which is the media converting ultraviolet light or/and high energy photons into hydroxyl radical ( • OH) [33].Meanwhile, it is believed that hydroxyl radical is one of the most strong oxidability matters.On the other hand, too much coating of TiO 2 film would not only increase the barrier of visible light to Pr:Y 2 SiO 5 nanomaterials, which weakens the intensity of incident light, but also decrease the proportion of Pr:Y 2 SiO 5 in the composite, in which the conversion of visible light to ultraviolet light is provided.As a result, the sample containing 5% TiO 2 presents a lower degradation rate comparing to the sample with 4% of TiO 2 .
The degradation curves of samples treated for 2 h are shown with an insert map at the right corner of Figure 7. Since an adsorbing balance of half an hour in a dark environment has been pretreated before the photocatalysis, the degradation data mainly resulted from the adsorbing capability of the samples, which are highly in accordance with the coating content of TiO 2 , as shown in Figure 7.The photodegradation performances of the samples in the next 1 h are highly affected by the adsorbing results which have the same tendency in the first half an hour.However, as the treatment time goes by, the photodegradation is dominant gradually which presents good evidence on the function of the upconversion materials in visible light.
The visible light excited photodegradation performances of TiO 2 @Pr:Y 2 SiO 5 before and after being loaded on GFF have been compared with the same content of composite catalyst, as shown in Figure 8.The sample TPY@GFF shows lower removing rates of pollutant than that of TiO 2 @Pr:Y 2 SiO 5 powder at all the treatment time except for at the first 2 hours which is attributed to the adsorption effect of GFF.The lower removing performances of TPY@GFF could be resulting from the embedding of the catalyst particles into the interspace of glass fiber filter and, as a result, lower contact surface of catalyst particles with pollutants than that of TiO 2 @Pr:Y 2 SiO 5 particles used directly.
Before the mineralization of those macromolecular compounds or refractory organics, such as benzene and its compounds, several degradation steps are needed and intermediate products would normally be produced.Total organic carbon (TOC), which has been tested by using a Shimadzu TOC analyzer (TOC-V CPH ) in this paper, is always used to characterize the organic carbon content and naturally the mineralization of organic pollutants in wastewater.Since the forming of the intermediates during the degradation of nitrobenzene, there always have been gaps between the TOC removing rates and the degradation rates which are tested by using the ultraviolet absorption testing method.However, these intermediates are still a kind of pollutant which can be tested by TOC.Therefore, the degradation rates are always a little bit higher than the removing rates of TOC, as shown in Figure 8.An interesting phenomenon is that more than 92% of nitrobenzene has been degraded and mineralized after 12 hours of photodegradation under visible light.When the treatment time comes to 24 hours, the photodegradation rate and mineralization rate are up to 98.9% and 98.5%, respectively.This suggested that the asprepared upconversion materials show excellent photocatalysis on the degradation of nitrobenzene under the exciting of visible light.
The best performance sample TPY@GFF, where the TiO 2 coating content is 4% in TPY material, has been used to study its reusability performance.The degradation time of each recycle degradation process has been set to 12 hours since the photodegradation rates are not increasing significantly after that, as shown in Figure 8.Therefore, after lasting for 12 hours of photodegradation, TPY@GFF samples are pulled out and heat-treated by drying at 80 ∘ C for 1 h and calcining at 500 ∘ C for 2 h sequentially to get rid of the impurities and renew the samples.The recycle degradation performances of TPY@GFF after 6 times of reusing are shown in Figure 9.As the reusing times increase, the degradation rates are getting down gradually from 95.0% of the first time to 75.9% of the sixth time.It is well known that, by the reusing times, the active sites of the catalysts would be decreasing inevitably and the catalysts particles loaded on the carrier would be dropping down more or less, no matter how advanced or careful the renew process is.However, the photodegradation rate is still more than 90% at the 3rd time and up to 75.9% even after reusing for 6 times, as shown in Figure 9.It suggests that the TPY@GFF materials have excellent visible light excited photodegradation, as well as good reusability performances.

Conclusions
Titanium dioxide TiO 2 coated Pr:Y 2 SiO 5 upconversion materials, TiO 2 @Pr:Y 2 SiO 5 , have been prepared by using a solgel method.The composite upconversion materials have been loaded on the glass fiber filter by using a dip-coating method.The luminescence intensities of the composite TiO 2 @Pr:Y 2 SiO 5 are decreased sharply by the coating contents of TiO 2 nanofilm, which could be attributed to the adsorbing of the luminescence emitted by the upconversion material Pr:Y 2 SiO 5 by the TiO 2 coating film in situ.As a result, the TiO 2 @Pr:Y 2 SiO 5 composite with 4% TiO 2 , which presents low luminescence intensity apparently, shows high efficiency on the photodegradation of nitrobenzene wastewater.Though a little bit lower photodegradation than that of TiO 2 @Pr:Y 2 SiO 5 powder used directly, the composite upconversion materials loaded on glass fiber filter, namely, TPY@GFF, show excellent reusability performances on the photodegradation of nitrobenzene and present 75.9% degradation rate by the treatment time of 12 h even after having been reused for 6 times.

Figure 6 :
Figure 6: Emission spectra of the samples.

Figure 8 :
Figure 8: The degradation rates and TOC removing rates of nitrobenzene.