Chaofei Fei,Dan Li,Xian Mao,Yu Guo,Wenheng Jing*

State Key Laboratory of Materials-Oriented Chemical Engineering,Nanjing Tech University,Nanjing 210009,China

Keywords:Ordered mesoporous MnOx Catalytic ozonation Pseudo- first-order reaction

A B S T R A C T In this account,highly ordered mesoporous MnOx/TiO2composite catalysts with efficient catalytic ozonation of phenol degradation were synthesized by the sol–gel method.The surface morphology and properties of the catalysts were characterized by several analytical methods,including SEM,TEM,BET,XRD,FTIR,and XPS.Interestingly,Mn doping was found to improve the degree of order,and the ordered mesoporous structure was optimized at 3%doping.Meanwhile,MnOxwas highly dispersed in the ordered mesoporous materials to yield good catalytic ozonation performance.Phenol could completely be degraded in 20 min and mineralized at 79%in 60 min.Thus,the catalyst greatly improved the efficiency of degradation and mineralization of phenol when compared to single O3or O3+TiO2.Finally,the reaction mechanism of the catalyst was discussed and found to conform to pseudo- first-order reaction dynamics.

1.Introduction

In recent decades,heterogeneous catalytic ozonation was increasingly explored as a promising technology for degradation and mineralization of refractory organic pollutants[1–3].For these processes,heterogeneous catalysts are crucial for promoting the generation of hydroxyl radicals[4,5].On the other hand,the structure and surface properties of the catalysts directly influence the active sites and the catalysis performance.Therefore,significant efforts have been devoted to modifying the structure and surface properties of catalysts.

Mesoporous materials,such as MCM-41,MCM-48 and SBA-15 have received considerable attention as metal oxide catalysts for catalytic ozonation[6–8].This is particularly due to their unique crystal structures,high specific surface areas,elevated pore volumes,and tunable pore size.The highly-ordered pore structure could enhance metal oxide dispersion and the interfacial electron transfer.Among mesoporous materials,highly-ordered TiO2was found promising because of its high surface area,energy band structure,and elevated conductivity[9,10].From thesynthetic viewpoint,highly ordered mesoporous materials were successfully prepared by introducing Mginto the alumina lattice framework[11].The addition of Ce during the synthesis of cubic mesoporous titania thin films significantly increases the degree of me so structural order[12].However,though TiO2mesoporous materials with various structures have been developed,the preparation of longrange ordered porous structures is still challenging.

Low-cost manganese oxides(MnOx)have been suggested as active components,which could accelerate the catalytic ozonation process.The hydroxyl radicals produced by the reaction of ozone on manganese oxide surface could improve the removal efficiency of refractory compounds.During this process,surface hydroxyl groups present on the catalyst play a key role.On the other hand,the preparation method and pH of the solution significantly influence the catalyst efficiency.Metal oxides may generate hydroxyl groups through dissociation or protonation in solution. It has been reported that positively charged surfaces of Mn/γ-Al2O3greatly influence the catalytic ozonation of phenol[13].However,some reports suggested that negatively charged surface groups contribute to ozone decomposition[14–16].In addition,multivalent manganese oxidation reaction occurs in ozone effect,with loss of electrons.This results in a reduction of active free radicals and high valence manganese oxide to complete the redox reactions and favors the catalytic ozonation.Enormous efforts have been devoted to investigating the catalytic mechanism of MnOxbut the outcomes are still ambiguous and sometimes controversial.

The traditional methods used for the synthesis of mesoporous materials include hydrothermal/solvothermal, microwave, and sonochemical processes. In the present work, a one-step sol–gel method combined with evaporation-induced self-assembly process was adopted for synthesis of the catalysts.If compared to the above-mentioned methods,the sol–gel process is easy to operate coupled with mild operating conditions and simpler experimental equipment[17–20].Manganese oxide was uniformly dispersed on TiO2framework to form new ordered mesoporous Mn–Ti composite transition metal oxides[21,22].The morphology and surface properties of the catalysts were analyzed by various characterization techniques,and phenol was selected to evaluate the catalytic ozonation activity of the resulting catalysts.

2.Experimental

2.1.Chemicals and materials

Phenol,ammonium chloride,ammonia,potassium ferricyanide and 4-aminoantipyrine,tert-butanol were supplied by the Sinopharm Company,Shanghai.Titanium(IV)tetraethoxide, hydrochloric acid,P123(EO20PO70EO20,Ma=5800),butyl alcohol,50 wt.%manganese nitrate solution,sodium hydroxide and hydrochloric acid were purchased from Aladdin,Shanghai.All the reagents were of analytical grade and used without further purification.

2.2.Preparation of catalysts

The catalysts were synthesized using the one-step sol–gel method.Precisely,HCl(8.4 ml)was added to 10 ml titanium tetraethoxide at 20°C under vigorous stirring.A solution of 3.76 g P123 and different quantities of manganese nitrate dissolved in 41 ml 1-butanol were then added to the HCl/Ti(OEt)4solution.The resulting Mn/Ti(mol/mol)molar ratios were as follows:0%,1%,3%,6%,and 10%.After 3 h,the sol solution was aged at 40°C and 80%RH for 12 h.Finally,the aged gel was sintered at 350°C for 4 h to obtain ordered mesoporous catalytic materials.

2.3.Characterization of catalysts

The wide-angle X-ray diffraction(WAXRD)patterns were measured by a Bruker D8 Advance X-ray diffractometer with Cu-Kα radiation.Small-angle X-ray diffraction(SAXRD)spectra of the samples were performed with an X-ray diffractometer(Ultima IV/PSK with Cu-Kα radiation λ=0.154 nm).The surface morphologies of the catalysts were viewed by a ZEISS LEO 982 GEMINI field emission electron microscope at the secondary electron mode,using the in-lens detector to improve resolution(CMA,Facultad de Ciencias Exactas y Naturales,UBA).The pore sizes of the catalysts were collected by a Philips EM 301 transmission electron microscope operated at 60 kV(CMA,Facultad de Ciencias Exactas y Naturales,UBA).The surface area and pore volumes were obtained from the N2adsorption/desorption isotherms conducted at 77 K on a Micromeritics Tristar II 3020 system.The specific surface area was calculated by the BET method and mesopore volume was derived from the adsorption isotherm according to the Barrett–Joyner–Halenda(BJH)model.X-ray photoelectron spectroscopy(XPS)data were collected on an AXIS-Ultra instrument using monochromatic Al Ka radiation(225W,15 mA,15 kV).Infrared spectral profiles were recorded on a Nicolet iS 50 Fourier transformation infrared(FTIR)spectrometer using KBr disks in the range of 4000–400 cm-1at room temperature.The zeta potentials of the catalyst suspensions were measured with aMalvern 3000 Zetasizer, using three consistent readings. The pH values of the solutions were evaluated by an acidimeter(pHs-3C,Shanghai,China),and concentrations of phenol were determined by a UV-1000 spectrophotometer(Gold S54).

2.4.Ozonation degradation of phenol by MnOx/TiO2catalyst

Typically,phenol solution(1L of 50mg·L-1)and 0.3g catalyst were added into a 1.5 L cylindrical reactor.Fig.1 shows a schematic illustration of the experimental setup.In experiments,ozone is produced through the air.The resulting ozone concentration is 12.43 mg·min-1.A propeller was employed to enhance the gas–liquid–solid three-phases mixing.All experiments are carried out at room temperature.After the reaction,the catalysts were filtered off and collected using a Millipore filter(pore size 0.22 μm).The residual ozone left in the solution was removed by adding 0.1 mol·L-1Na2S2O3solution.The concentration of phenol was measured by the chemical indication process.Ammonia buffer solutions containing ammonium chloride,4-aminoantipyrine and potassium ferricyanide were respectively added to the solution and absorbance was measured after 10 min at 510 nm using a UV–Vis spectrophotometer.The total organic carbon(TOC)of the solution was determined using a Phoenix 8000 TOC analyzer.

The catalytic efficiency of the heterogeneous catalytic ozonation highly depends on load of Mn(II)oxide in the catalyst and pH of the solution.Thus,the catalytic efficiency with variable Mn(II)oxide loading contents and pH values were examined.The removal efficiencies of phenol(RPhenol)and TOC(RTOC)were selected to evaluate the catalytic efficiency of MnOx/TiO2catalyst,which were calculated using Eqs.(1)and(2).

Fig.1.A schematic representation of the experimental setup.

where c[Phenol]0and c[TOC]0were the initial concentrations,and c[Phenol]tand c[TOC]twere the concentration of phenol and TOC at time t(mg·L-1),respectively.

3.Results and Discussion

The crystallinity of the resulting MnOx/TiO2catalysts was examined by XRD,and the results are shown in Fig.2a.It can be seen that the catalysts mainly contained TiO2crystalline phase due to the small load in MnO.The peaks at 25.3°,38.3°,48.2°,54.5°and 62.5°were assigned to typical anatase.However,it has to be kept in mind that the diffraction peak intensity of anatase phase decreased as MnO load increased.Furthermore,several studies have demonstrated that the degree of crystallinity could be suppressed using doping elements with larger atomic radius.In the present case study,Mn2+has an atomic radius of 0.067 nm,which is larger than that of Ti4+(Ti4+:0.0605 nm)[23,24].This may induce a distortion of the TiO2lattice as Ti4+was replaced by Mn2+,resulting in inhibition of the TiO2crystal growth.Additionally,no refraction peak attributable to Mn was observed in Fig.1a,indicating that Mn was highly dispersed on the TiO2surface.

Fig.2b shows the SAXRD curves of TiO2samples at different MnO contents.The presence of the peaks at 2θ =1.00°,1.24°and 1.57°at low angle range(2θ =0–5°)indicated the presence of an hexagonal mesophase[11,25].The blank TiO2materials presented no obvious diffraction peaks,suggesting a dissatisfactory pore order degree.In the presence of MnO,the pore order significantly improved,especially at MnO doping amounts of 3%.During the sintering process,the ordered pore structure was prone to collapse as a result of template removal or growth of the crystal.The addition of MnO suppressed the crystal growth of TiO2samples,retaining the ordered pore structure during the crystal growth.Nevertheless,as MnO doping amount further increased,the diffraction peak was weakened and no obvious diffraction peak occurred at doping amounts of 6%and 10%.The latter could decrease the stability of the sol,leading to weakening of the catalyst structure.

The surface morphologies of the materials were characterized by SEM,and the results are depicted in Fig.3a–b.It will be noted that both surfaces of prepared TiO2films appeared smooth.Furthermore,the TiO2film containing 3%Mn presented an ordered stripy pore structure.To gain further understanding of TiO2membrane microstructure,TEM was used and the results are shown in Fig.3c–d.The prepared TiO2membrane clearly presented an unordered pore structure.In the presence of Mn,a typical 2D hexagonal pore structure was observed.These data indicated that the doped Mn consolidated the pore to maintain the long ordered pore structure when the template was removed,consistent with SAXRD and SEM results.

Fig.3.(a,b)SEM and(c,d)TEM of the MnOx/TiO2catalyst with different Mn loading amounts(a,c-0%,b,d-3%),(e)EDS elemental distribution map of O(red),Ti(orange),and Mn(yellow).

The sample 3%-MnOx/TiO2scanned by EDS(Table 1)estimated the Mn/Ti(mol/mol)ratio to 0.0318,consistent with 3%.This further indicated that the active center elements were all uniformly dispersed on the support surface.The elemental distribution map for O(red),Ti(orange)andMn(yellow)are listed in Fig.3e.The Mnelement appeared more evenly distributed in the material to form complex metal oxides.

Table 2 illustrates the microstructural properties of the TiO2samples.In the presence of Mn,both the BET specific surface area and pore volume significantly improved.The 3%-MnOx/TiO2presented the highest BET specific surface area of 280.2 m2·g-1and pore volume of 0.434 cm3·g-1.These values were almost 1.5-fold higher than those of pure TiO2.As Mn loading increased,both the specific surface area and pore volume decreased.This was mainly attributed to blockage of the mesopores and channels in the TiO2support.In addition,the radii of Mn2+and Ti4+were similar,in the process of catalyst synthesis,Mn2+is combined with Ti4+instead of the combination between partial Ti4+and Ti4+.Channel structures are formed by self-assembly,the average pore size changes little.Hence,it could be concluded that Mn loading changed the microstructure of TiO2.This was consistent with data of Fig.3b,showing that doping with 3%Mn raised the order structure of the material to the maximum.This,in turn,was beneficial to raise the specific surface area.

Fig.2.(a)The wide and(b)low angle XRD patterns of the different catalysts.

Table 1Percentage of elements in the catalyst

Table 2BET surface area,pore volume and pore diameter of catalysts with different Mn loadings

Fig.4.Plot of zeta potential as a function of pH for different catalyst suspensions.

The active sites containing the surface hydroxyl on the catalyst splay an important role in the catalytic ozonation process.The pH of the solution also directly influences the formation of the surface hydroxyl groups.At pH deviated from the point of zero charge(PZC),the catalyst's surface could either be protonated or deprotonated to yield a positively or negatively charged surface.Therefore,the zeta potential of different catalysts as a function of pH was investigated to figure out the influence on the catalysts doped with Mn,and the results are gathered in Fig.4.Apparently,the addition of MnOxdecreased the PZC.The PZC values of the catalyst were recorded as～6.1,5.6,4.7,3.8 or even lower when MnOxconcentration varied from 0 to 10%.Published values of PZC of titanium dioxide and manganese dioxide were 6.2 and 2[26,27].Thus,the addition of manganese oxide greatly changed the surface properties of titanium dioxide.

Fig.5a represents the degradation of phenol process for both single ozonation and catalytic ozonation performed with different catalysts at pH=7.Only 85%of phenol was removed during single ozonation after 60 min.The addition of the catalyst promoted the degradation of phenol.In the presence of pure TiO2,phenol was completely removed after 60 min while the time was reduced to 40 min using MnOx/TiO2,at the exception of 10%-MnOx/TiO2.Fig.6a clearly showed that at relatively low MnOxamounts(≤3%),the catalytic ozonation activities gradually increased with loading amount.The catalytic ozonation activity of the 3%-MnOx/TiO2sample reached a maximum,demonstrating the optimal load ratio of MnOxto 3%.Further increase in loading amount of MnOxgradually reduced the catalytic ozonation activities.MnOx/TiO2was then concluded to be an effective catalyst for ozone.

In general,the catalytic ozonation is highly efficient in mineralization of organic pollutants.The removal efficiencies of TOC during the process of phenol degradation are shown in Fig.5b.In single ozonation,only19%of TOC was removed after 60 min.This lower removal efficiency of TOC compared to phenol indicated that phenol was not completely oxidized to CO2and H2O under the ozone environment.In the presence of TiO2particles,the removal efficiency slightly improved to reach 21%but remarkably increased when MnOx/TiO2was added to reach 79%with 3%-MnOx/TiO2.On the other hand,excess manganese(6%and 10%)appeared to reduce the removal rate of TOC.This may be due to the large amounts of Mn doping that led to the enrichment of manganese oxide and collapse of the ordered structure.In turn,this caused the catalytic ozonation activity to decrease.These data demonstrated that insertion of 3%Mn in the titanium dioxide lattice played an important role in self-structure of the material.This led to formation of a Mn–Ti blend transition metal oxide ordered mesoporous molecular sieve.The optimized catalyst accelerated the degradation and mineralization of phenol when compared to single ozonation and disordered mesoporous TiO2.

Fig.5.(a)Degradation of phenol and(b)TOC removal during ozonation and catalytic ozonation of phenol at pH 7.

Fig.6.Influence of t-BuOH on(a)the degradation efficiency of phenol and(b)TOC removal during ozonation and catalytic ozonation at pH 7.

To further explore the surface properties of the MnOx/TiO2catalyst,FTIR analyses were performed and the results are shown in Fig.7.It can be seen that blank sample contained two OH absorption bands.The band at 3600 cm-1was attributed to stretching of OH groups of adsorbed water molecules and the peak at 1622 cm-1corresponded to OH bending vibration of adsorbed water[5].In the presence of MnOx,a new absorption peak at 1384 cm-1appeared which was assigned to OH deformation vibrations of hydrated MnOx.It will be noted that introduction of MnOxslightly decreased the intensity of Ti--O bond vibration,indicating the formation of a Ti--O--Mn band.The 3%-MnOx–TiO2catalyst was also characterized by FTIR after a catalytic ozonation,and the results are gathered in Fig.7b.After ozone aeration,similar absorption bands appeared but the characteristic peak intensity of MnO-H and H2O decreased.This confirmed the surface hydroxyl groups and chemical adsorption of water as the active sites participating in the reaction.

The XPS characterization of the surface chemical species of catalysts is shown in Fig.8 and the data are listed in Table 3.Two peaks in the O1s spectra can be observed.The peak at lower binding energy(529.21 eV)was attributed to lattice oxygen(O(-II),denoted Oa),and the peak at 531.46,532.30eV was ascribed to adsorbed oxygen and surface hydroxyl species(denoted Ob).The peak at the binding energy of 532.01 eV was associated with the presence of oxygen in Ti--O--Mn(denoted Oc)[28–30].It also showed that Mn was successfully embedded into the titanium oxide lattice,consistent with the XRD data.For fresh catalysts,the Obpeak area first increased then decreased as Mn content rose.There was a huge leap in the amount of Obfor the 1%Mn TiO2sample,which reached a maximum value of 72.88%when Mn was 3%.However,a further increase in the doping amount reduced the surface hydroxyl groups.

The samples were also analyzed using XPS,and the spectra of Mn 2p are shown in Table3.The peaks at640.78and641.57eV were attributed to the presence of Mn(III),and the peak at 646.88 eV was assigned to Mn(IV)[31].After catalytic ozonation,the relative contents of Oaincreased from 16.00%to 39.21%,Obreduced from 72.88%to 56.34%,and Ocdeclined from 11.12%to4.45%.These results indicated that lattice oxygen,surface hydroxyl and adsorbed water were all involved in the catalytic process[32–34].The XPS results of Mn2p in the 3%-MnOx–TiO2partly changed after the catalytic ozonation.Furthermore,the relative contents of Mn(III)decreased from 74.12%to 62.54%,and those of Mn(IV)changed to 37.46%.These findings revealed that Mn at+3 valent state before ozone oxidation reaction was partly oxidized to+4 valence state due to the strong oxidation effect of ozone.The ozone molecules generated further active free radicals by obtaining electrons.Overall,this demonstrated that the oxidation state of Mn gradually shifted from lower to higher oxidation states,suggesting the occurrence of Mn(III)/Mn(IV)redox reaction during the catalytic ozonation process.

In general, tert-butanol is considered as a very strong radical scavenger,which could react with the hydroxyl radicals to generate inert intermediates and inhibit the chain reaction of radicals. The reaction rate constants of tert-butanol with ozone and hydroxyl radicals were reported as3.0 × 10-3L·mol-1·s-1and 5.0 × 108L·mol-1·s-1,respectively[35].According to the above results,the phenol solution at pH=7 was employed to study the reaction mechanism and Fig. 6a illustrates the removal efficiencies of phenol with and without tert-butanol.In the absence of tert-butanol,the degradation efficiencies of phenol were determined as 85%and 100%after 60 min for single and catalytic ozonations,respectively.Nevertheless,the degradation efficiency of single ozonation was less inhibited in the presence of 10 mmol·L-1tert-butanol.Compared with ozonation alone,the degradation efficiency of the catalytic ozonation significantly declined.One possible reason was that the ozone molecules present on the surface active sites of the catalyst underwent decomposition to produce active free hydroxyl radicals(·OH)to participate in the degradation of phenol.The addition of tertbutanol in the solution induced a reaction with the hydroxyl radicals to slow down the catalytic ozonation reaction rate.Hence,this reaction was probably governed by hydroxyl radical mechanisms.

Fig.7.FTIR spectra of different samples(a)before and(b)after the catalytic reaction.

Fig.6b shows the mineralization of phenol with and without tertbutanol.The TOC removal rate slightly reduced from 16%to 15%in the presence of tert-butanol in single ozonation.In neutral and acid media,ozone is mainly present at molecular state in solution to directly react with organic compound. Therefore, tert-butanol showed a little effect on ozonation alone.For catalytic ozonation,the ozone molecules were adsorbed onto the surface of the catalyst to rapidly generate free radicals which quickly combine with tert-butanol rather than organic molecules.This,in turn,resulted in a serious decline of the TOC removal efficiency.

In addition to nature of the catalysts, the solution pH significantly influences the efficiency of the catalytic ozonation reaction. As mentioned above,the solution pH mainly influences the formation of the surface hydroxide groups.It has been demonstrated that hydroxide(OH-)groups could initiate a chain reaction of ozone decomposition,and elevated amounts of hydroxyl radicals(·OH)are generated at higher pH[36].Therefore,a faster and higher degradation efficiency could be obtained in alkaline environment for single ozonation[37].For heterogeneous catalytic ozonation,charge properties of the catalyst may play an important role in ozone decomposition dependently of the solution pH.

Fig.8.XPS spectra of O 1s and Mn 2p of the samples.

Fig.8(continued).

Therefore,the decomposition of phenol solution in different pH environments(3,5,7,9)was investigated, and the results are summarized in Fig. 9. It can be seen that the degradation efficiency of phenol in single ozonation increased from 72%to90%after 60 min as the pH rose from 3 to 7(Fig.9a),and the removal efficiency of TOC increased from 16.8%to 20.9%(Fig.9c).For pH above 9,almost all phenol was degraded after 40 min,and the TOC removal efficiency further increased to 29.5%.A similar phenomenon was observed in Fig. 9b,where the degradation efficiency enhanced at pH smaller than 7. However, the mineralization efficiency decreased at pH=9.The catalyst surface covered with hydroxyl groups would be protonated or deprotonated depending on the pH value of the solution(lower or higher than pHpzc).Several studies demonstrated that deprotonated or neutral surface hydroxyl groups had a strong reactivity towards ozone molecules[5,38,39].The pHpzcof the 3%-MnOx/TiO2was determined as 4.7.Therefore,the surface hydroxyl groups were present at the deprotonated states for pH＞4.7.Xing et al.reported that a negatively charged surface had a strong reactivity towards ozone[5,14].The·OH radicals generated from the reaction of deprotonated surface hydroxyl groups(≡MnO-)and O3could be expressed through Eqs.(3)–(5)[14].

When pH less than or equal to 7,phenol with a state of positive ion adsorption to the catalyst surface, and with the rise of pH, catalytic reaction rate is faster and faster.When pH ＞ 7, phenol with anion state exists in solution gradually.So when pH=9,phenol anion and catalyst surface repel each other,and it is not conducive to the adsorption and reduces the catalytic reaction rate.In general,the increase in the initialpH leads to larger densities of the surface hydroxyl groups,which further produces·OH radicals.Unfortunately,the degradation efficiency declined as pH further increased.Besides,the interfacial electron transfer is involved in the catalytic decomposition of ozone,where electrons from the surface of Mn(III)decomposed ozone to hydroxyl radicals,leading to elevated activities towards phenol mineralization.Therefore,the dispersion and oxidation states of supported MnOxwere crucial factors in obtaining higher catalytic ozonation efficiencies.

Table 3O 1s and Mn 2p binding energies(C 1s at 284.6 eV)

In addition,we studied the solubility of catalyst at different pH.From Fig.9d,we can find that Mn dissolution loss falls down from 1.33 to 0.41 mg·L-1when pH ranges from 3 to 9.The results illustrated that Mn dissolution loss is low in alkaline conditions,and it is beneficial to recycling the catalysts.

The mechanism of hydroxyl radical production is displayed in Fig.10.The interaction between the ozone molecules and surface active site was also present. In addition, a series of intermediate reactions composed of types(3)–(5)and multivalent manganese oxide by electron transfer to produce hydroxyl radicals for subsequent catalytic degradation of phenol were displayed.

The degradation of organic compounds is often divided into two reaction pathways:direct ozonation and indirect ozonation[40].For direct ozonation,organic compounds are directly attacked by O3.However,contact with organic matter to generate hydroxyl radicals from the catalytic occurs in indirect ozonation[41].It has been demonstrated that the kinetics of the catalytic ozonation was governed by second-order or pseudo- first-order mechanisms[42,43].In this study,the degradation rate of phenol by MnOx/TiO2catalytic ozonation could be written as follows:

where kO3and k·OHare the constants for ozone and ·OH reacting with phenol,respectively.Thus,Eq.(6)could be transformed into Eq.(7)through:

Fig.9.Degradation of phenol of(a)single ozonation,(b)catalytic ozonation,and(c)TOC removal during ozonation and catalytic ozonation of phenol at different pHs,(d)Mn leaching at different pHs.

Fig.10.Mechanism of hydroxyl radical production.

Fig.11.The kinetic of the reaction rate of phenol degradation(a)with different catalysts at pH 7,(b)single ozonation and(c)catalytic ozonation at different pHs.

By defining kr=kO3[O3]+k·OH[·OH],Eq.(7)could be trans formed to Eq.(8):

In order to explore the influence of the doping amount on the reaction kinetics of phenol degradation,different catalysts at pH 7 were evaluated and the results are represented in Fig.11a and Table 4.The correlation degree of all the pseudo- first order reactions appeared very high,consistent with requirements for fitting data.The slopes of fitted lines represent the reaction kinetic constants.For O3,the kinetic constant was determined as 0.03614 min-1,and the linear fit was 0.98958.When compared to O3,the accelerated reaction rate for 0%-MnOx–TiO2catalyst may be due to phenol adsorption onto the surface.As Mn doping increased,the kinetic constant rose.This was related to MnOxas the active center of the catalytic reaction,which accelerated the degradation of phenol.The maximum kinetic constant was reached at doping amount of 3%,then started to reduce above this percentage.This could be related to the formation of most ordered pore structure at 3%doping.It is well known that ordered channel is favorable for mass transfer and diffusion,and regular pore structure may provide a more specific surface area toenhance the catalytic reaction.Excess doping decreased the order structure,which would affect mass transfer,adsorption,diffusion,and further reduce the catalytic efficiency.

Table 4Pseudo- first-order rate constants of phenol degradation in ozonation at pH 7

The influence of pH on the reaction kinetic rate of phenol degradation was also evaluated for both single ozonation and catalytic ozonation processes.As depicted in Fig.11b–c,both processes were governed by pseudo- first order reactions.Table 5 listed the kinetic constants and fitting degrees.For single ozonation,the degradation rate of phenol gradually increased as pH rose.In alkaline conditions,the degradation rate remarkably improved due to the reaction of ozone molecules with hydroxyl ions generated hydroxyl radicals.This was beneficial for degradation of phenol.For 3%-MnOx/TiO2,the reaction rate constant achieved its maximum at pH 7,which was 5.5-fold higher than single ozonation.In addition,shifting pH from neutral to alkaline caused the rate constant to reduce when compared with single ozonation that only increased by 1.5-fold.

Table 5Pseudo- first-order rate constants of phenol degradation in single and catalytic ozonations

4.Conclusions

Titanium dioxide supported manganese oxide catalysts were successfully synthesized by the sol–gelmethod. The addition of manganese ions improved the mesoporous titania structural order.The analyses of the materials showed that their properties were improved,and the specific surface area and surface active sites were increased.A 3%Mn doping was found to be the best doping amount.Other data suggested that addition of the catalyst accelerated the degradation and mineralization of phenol.The reaction kinetics revealed that the catalytic ozonation reaction at pH=7 yielded the largest reaction kinetic constant,faster than values obtained at other pHs or ozonation performed alone at different pHs.The addition of tert-butanol to the phenol solution at pH=7 showed that the catalytic ozonation was governed by the hydroxyl radical reaction mechanism.