Catalytic Wet Oxidation of Phenol with Mn Ce-Based Oxide catalyst

Catalytic Wet Oxidation of Phenol with Mn-Ce-Based Oxide Catalysts:Impact of Reactive Adsorption on TOC RemovalM.Abecassis-Wolfovich,†ndau,*,†A.Brenner,‡and M.Herskowitz†Chemical Engineering Department,The Blechner Center for Industrial Catalysis and Process Development,and Biotechnology and Environmental Engineering,Ben-Gurion University of the Negev,Beer-Sheva84105,IsraelCatalytic wet oxidation of phenol solutions at low temperatures of80-130°C and space velocitiesof1-100h-1using Mn-Ce catalysts was studied with an emphasis on the reactive adsorptionmechanism and total organic carbon(TOC)removal.Eight catalysts(Mn/Ce)6:4)were activatedunder different conditions and promoted with alkali metals(K,Cs)or noble metals(Pt,Ru).The compositions and physical properties of all catalysts were measured.Preliminary runs wereconducted in a batch reactor,but most experiments were carried out in a continuous-flow trickle-bed reactor.Catalysts containing mixed Mn3O4-CeO2phases pure and promoted with alkalimetals displayed a higher activity and a higher adsorption of organic deposits on their surface.Noble metals had little effect on process performance.The adsorption capacity of the catalystswas found to be considerably higher than that reported for activated carbon.Furthermore,complete regeneration of a catalyst in three consecutive tests was demonstrated under relativelylow temperature and with no loss of activity.The selectivity toward reactive adsorption washighest on Mn-Ce-Cs catalysts.Low space velocity yielded essentially complete adsorption ofphenol,resulting in deposits on the catalyst surface.The conversion of phenol to water-solubleoxygenates was found to increase water toxicity.The catalytic reactive adsorption-regenerationprocess should become an attractive treatment method for phenol solutions and other complexwaste streams.IntroductionCatalytic wet oxidation(CWO)with heterogeneous catalysts is a method for the treatment of dilute aqueous waste streams containing a variety of organic pollutants such as phenols.1,2It can be used to mineralize organic contaminants to CO2and H2O or to convert them into nontoxic and biodegradable products.CWO has also become an alternative to the traditional treatment of wastewater with activated carbon,in a periodic adsorp-tion-regeneration process.3,4CWO needs to be con-ducted at relatively low temperature to exclude water evaporation and condensation,short space-time to save reactor volume,and with a high-recovery regeneration procedure to ensure economic feasibility.Many catalysts have been tested in CWO of phenols: Pt;Pd;Ru/C;5CuO-ZnO/Al2O3;6,7Ru/CeO2;8Pt/CeO2;9 and bulk manganese-cerium oxide composites,pure10-12 or doped with Pt,13Pt-Ag,14or K.15,16Only composites based on manganese-cerium oxides have displayed the ability to remove phenol efficiently at low temperatures of e100°C.At low temperatures,Mn-Ce-based cata-lysts convert phenol to polymeric carbonaceous deposits produced on cerium-related sites that cause substantial catalyst deactivation due to surface blocking.10,12,17The amount of carbon on a spent Mn-Ce catalyst could reach about23wt%.11Introduction of Pt decreases the amount of carbonaceous deposits,13whereas promoting the Mn-Ce catalyst with potassium achieves the op-posite effect.15Pintar and Levec7measured the formation of poly-meric deposits on the catalyst surface during phenol CWO on CuO-ZnO/Al2O3.They assumed that poly-merization was promoted by the partial oxidation of phenol to aldehydes followed by condensation with phenol.Such polymerization usually occurs on basic catalysts.18It is therefore consistent with promotion of carbon deposition on Mn-Ce composites on cerium-related sites,10,12,17which are more basic than the surface of manganese oxide.19,20The polymeric deposi-tion on the catalyst surface can be considered as the “reactive adsorption”of phenol caused by the catalytic material.Its contribution to TOC removal in phenol CWO reached50-70%.7,13Mn-Ce-based catalysts were tested in phenol CWO at low temperature only in a batch reactor,10,11,13although it was already postulated that the process should be more efficient in a trickle-bed reactor because of the higher solid-to-liquid ratio.7,21 No information has been published on the regeneration of Mn-Ce-based catalysts after deactivation during the CWO of phenol.The scope of the present study was to demonstrate a controlled reactive adsorption mechanism as a means to improving TOC removal in the CWO of phenol with Mn-Ce-based catalysts in a fixed-bed reactor.Oxidative regeneration procedure to recover the initial catalyst activity was also developed.The information obtained was compared to the conventional adsorptive removal of phenol from wastewater by activated carbon. Experimental WorkCatalyst Preparation.Two samples of manganese and cerium oxide composite(atomic ratio Mn/Ce)6:4) were prepared by coprecipitation from mixed aqueous*To whom correspondence should be addressed.E-mail: mlandau@bgumail.bgu.ac.il.†The Blechner Center for Industrial Catalysis and Process Development,Chemical Engineering Department.‡Biotechnology and Environmental Engineering.5089Ind.Eng.Chem.Res.2004,43,5089-509710.1021/ie049756n CCC:$27.50©2004American Chemical SocietyPublished on Web07/13/2004solutions of manganese(II)chloride(MnCl2‚4H2O,Sigma Chemical Co.)and cerium(III)chloride(CeCl3‚7H2O, Sigma Chemical Co.)as described by Imamura et al.22 One hundred milliliters of this solution was poured into 200mL of3M aqueous sodium hydroxide(NaOH,97%, Aldrich Chemicals),and the resulting precipitate was separated by filtration,washed,dried in air at100°C for16h,and then calcined at350°C for3h under a vacuum of85mbar(Mn-Ce-1)or in air(Mn-Ce-2). Addition of potassium or cesium to these Mn-Ce catalysts was done by incipient wetness impregnation that was performed after drying the Mn-Ce precipitate at100°C.The samples were impregnated with aqueous solutions of potassium nitrate(KNO3,Aldrich Chemi-cals)or cesium nitrate(CsNO3,Aldrich Chemicals)to gain4wt%of the alkaline metal.The samples were evacuated at350°C for4h(Mn-Ce-K-1,Mn-Ce-Cs-1)or calcined in air under the same conditions(Mn-Ce-K-2,Mn-Ce-Cs-2).Addition of platinum or ruthe-nium was performed by incipient wetness impregnationof Mn-Ce catalyst after treatment in a vacuum at350°C for3h with5mg/L aqueous solutions of H2PtCl6or H2RuCl6(Aldrich Chemicals)to gain3wt%of the noble metal.After the samples had been dried at100°C,they were activated in hydrogen for3h at350°C(Mn-Ce-Pt and Mn-Ce-Ru).Catalyst Characterization.The chemical composi-tions of the catalysts(weight percentage,average of five measurements at different points of the solid)were measured by energy-dispersive X-ray spectroscopy(JEOL JEM5600scanning electron microscope).Surface areas and pore volumes were obtained from N2adsorption-desorption isotherms using the conventional BET and BJH methods.The calcined samples were outgassed under vacuum at250°C.Isotherms were measured at liquid nitrogen temperature with a NOVA-2000(Quan-tachrome,Version7.01)instrument.The phase compo-sitions of the catalysts were tested by X-ray diffraction (XRD).The XRD patterns were collected on a Philips diffractometer PW1050/70(Cu K R radiation)with a graphite monochromator at diffracted beam.Data were recorded at a0.02°step size for2s at each step.The peak positions and the instrumental peak broadening ( )were determined by fitting each diffraction peak by means of APD computer software.The crystal domain size was determined using the Scherrer equation where K)1.000, )0.1°,λ)0.154nm,and B is the peak broadening at2θ)30-45°for different manga-nese oxides and at2θ)48°for cerium oxide.Temper-ature-programmed reduction(TPR)and temperature-programmed oxidation(TPO)experiments were carried out in an AMI-100Catalyst Characterization System (Zeton-Altamira)equipped with a mass spectrometer for outlet component identification(Ametek1000).Catalyst (0.2g)was loaded and treated in10vol%H2-Ar,25 cm3/min(TPR),or5vol%O2-He,25cm3/min(TPO), as the temperature was increased from ambient to420°C at5°C/min.Catalyst Testing Procedures.The wet oxidation of phenol(Sigma Chemical Co.)was carried out in batch and fixed-bed reactors.The preliminarily tests were performed in a stirred autoclave reactor[steady state (SS),300mL,Bu¨chi]at100°C,under a10atm oxygen pressure,with a stirring speed of800rpm,a catalyst loading of2g/L,and a phenol concentration of0.2g/L. Catalyst(0.3g in a powder form to avoid any diffusion limitations)and110mL of distilled water were fed to the reactor,which was pressurized with oxygen to9 atm,and the mixture was then heated under agitation using an oil bath.After the desired reaction temperature had been attained,25mL of concentrated phenol solution(1.2g/L)was loaded into the reactor using an Eldex piston pump.One-milliliter samples were taken periodically for HPLC analysis(GBC,LC1205instru-ment,Zorbax ODS-5C18reverse-phase column,mobile phase acetonitrile/distilled water mixture(8:2vol ratio), 1cm3/min flow rate,UV spectrophotometer for detec-tion,λ)260nm)to determine the phenol concentration. Selected catalysts were tested in a continuous mode using a fixed-bed reactor(SS,20-mm i.d.,24-cm length). CWO was performed at80-130°C with10bar of oxygen pressure and a liquid hourly space velocity (LHSV)of1-100h-1in a fixed-bed reactor rig(Figure 1).The reactor was heated with an electric coil con-trolled by a Eurotherm PID controller.A K-type ther-mocouple was positioned in the center of the catalyst bed.Two independently controlled heating zones kept the axial temperature gradient in the reactor to less than5°C.The oxygen was fed by a Brooks mass controller to a preheater at a rate of25-100mL min-1 (STP conditions)before entering the reactor at the selected temperature.The tested solution containing1 g/L phenol(doubly distilled water)was fed to the reactor by a high-pressure metering pump(SSI HPLC series II isocratic)at20-52mL/h and reaction pressure. Catalyst pellets(0.3-11g,0.4-0.5mm fraction)were diluted with quartz particles of the same size in order to keep a fixed volume of catalyst layer inside the isothermal zone.The liquid effluent was collected in periods of time and analyzed by HPLC to detect the residual phenol concentration.The liquid outlet stream was also analyzed for TOC(total organic carbon)using an Apollo9000HS model TOC combustion analyzer (Tekmar Dohmann)equipped with a nondispersive infrared(NDIR)detector.Testing of phenol oxidation at100°C in a reactor loaded only with glass Raching rings showed that the phenol conversion was below1%. The phenol conversion(X PhOH,%)and TOC conversionl)Kλ[(B2- 2)0.5cos(2θ/2)]Figure1.Fixed-bed reactor rig for phenol oxidation:1,oxygencylinder;2,flowmeter;3,preheater;4,fixed-bed reactor;5,temperature controller;6,separator and liquid sampling cell;7,gas sampling;8,nitrogen cylinder;9,liquid pump;10,feed tank.5090Ind.Eng.Chem.Res.,Vol.43,No.17,2004(X TOC,%)were calculated asandwhere C PhOH is the phenol concentration,with the subscripts0and t representing the initial concentration for the batch reactor or the concentration at the inlet for the fixed-bed reactor(C PhOH,0)and the final concen-tration for a batch reactor or the concentration at the outlet for the fixed-bed reactor(C PhOH,t).C TOC is the TOC concentration measured at the fixed-bed reactor inlet or outlet.The reaction rate constants of phenol trans-formation in the batch reactor were calculated assuming first-order kinetics with respect to phenolTo examine the leaching of the metal ingredients,the effluent stream was also analyzed for the presence of Mn and Ce by an inductively coupled plasma(ICP) technique,using an Optima3000Perkin-Elmer ICPOES system.In addition,toxicity bioassays based on the relative inhibition of bioluminescence of marine bacteria were conducted.Toxicity was determined by the Vibrio fischeri bioluminescence assay method,using a model ToxAlert10luminometer(Merck K G a A,Darmstadt, Germany).The Vibrio fischeri bioluminescence assay has been successfully employed for the evaluation of toxicity in complex waste samples and has shown a high sensitivity to the presence of toxicants as well as a good correlation to other toxicity bioassays(Bulich et al.23).A working solution of luminescent bacteria was pre-pared by reconstituting a vial of freeze-dried Vibrio fischeri cells,purchased from Merck,using1mL of2% (w/v)NaCl.Luminescence of a tested solution and a control having the same amounts of reconstituted bacteria at pH6-8were conducted simultaneously. Toxicity was calculated as the percent decrease in luminescence of the tested solution,compared to the control,and is termed here the percent inhibition. Catalyst Regeneration.The amount of the carbon-aceous material adsorbed on the catalyst surface in continuous runs was calculated according to the time profile of CO2formed during oxidative regenerative treatment.In this treatment the spent catalyst was heated in the reactor under a50mL/min flow rate of oxygen and heating rate of3°C/min up to the required temperature and kept at this temperature until no carbon dioxide was detected at the reactor outlet.A gas sample was taken every few minutes from the exit pipe of the system and analyzed by GC(Gow-Mac580 instrument equipped with a TCD detector and a Pora-pack Q packed column;L)6ft,o.d.)1/8in.,i.d.)3 mm).Integration of the CO2concentration versus time data using Polymath5.1(Simpson method)yielded the total amount of CO2produced during oxidative regeneration. The corresponding mass of phenol on the surface (m PhOH,s)was calculated from this value and compared with the mass of reacted phenol(m PhOH,r).The ratio between those two values represents here the selectivitytoward reactive adsorption(SRA)and was calculatedasThe catalyst adsorption capacity(CAC)was calculatedaswhere m cat is the catalyst mass.When the regenerationwas performed after a small decline in phenol conver-sion(the start of visible deactivation),the ratio of phenolamount on the surface to the amount of catalyst wasdefined as the steady-state catalyst adsorption capacity(SS-CAC).When the regeneration was performed afterfull deactivation(X PhOH e30%),the ratio was defined as the total catalyst adsorption capacity(total CAC).SRA and CAC measurements and calculations wereperformed three times with an accuracy of5-10%. Results and DiscussionCatalyst Properties.Mn-Ce catalysts containingdifferent promoters(that affect polymeric deposition)were tested in a batch reactor.The catalysts testedincluded a mixed oxide with optimal CWO Mn/Ceatomic ratio of6:4established in previous investiga-tions.12,22This mixed oxide was modified with potas-sium,cesium,and noble metals(i.e.,platinum orruthenium)to examine its ability to accumulate car-bonaceous deposits during phenol CWO due to differentsurface basicity(K,Cs)and oxidation activity(Pt,Ru).The texture parameters of the different catalysts as wellas their phase compositions are listed in Table1.Thecatalysts were treated(1)in air or(2)under vacuumatmosphere,which caused a change in the structure ofmanganese oxide component from Mn5O8to Mn3O4.XRD diffractograms of samples Mn-Ce-1and Mn-Ce-2(Figure2)shows sharper peaks for Mn-Ce-1,indicatinga lower dispersion of Mn3O4compared to Mn5O8(crystalsizes of23-25and9-10nm,respectively)and similardispersion of the CeO2phase(4-5nm)(Table1).Therelatively high surface area of the mixed oxides(about100m2.g-1)is determined mostly by the CeO2phase,which displayed a much higher dispersion and com-prised about60%of the catalyst by weight.Taking intoaccount the theoretical densities of Mn5O8and CeO2of4.933and7.214g/cm3,respectively,the calculation ofcorresponding surface areas according to formula(where F is the density,d is the crystal diameter)gives 208m2/g for CeO2and122m2/g for Mn5O8.The weight ratio of these phases in the Mn-Ce catalyst treated in air is CeO2/Mn5O8≈1.5,and the measured surface area is128m2/g.This figure compared with the weighted contributions of the calculated surface areas of the oxide components,174m2/g,means that∼75%of the oxide crystal surface in the composite is accessible for reacting molecules.Evacuated Mn-Ce-1catalyst had a lower surface area relative to the air-treated Mn-Ce-2be-cause of the lower dispersion of the manganese oxideXPhOH (%))(1-C PhOH,t C PhOH,0)×100X TOC (%))(1-C TOC,outletCTOC,inlet)×100-ln(1-X PhOH100))ktSRA)mPhOH,smPhOH,rCAC)mPhOH,smcatSA)6000F dInd.Eng.Chem.Res.,Vol.43,No.17,20045091phase (23-25nm,Table 1).It is important to note that the final treatment affected only the manganese oxide phase and not the cerium oxide phase.Modification of Mn -Ce mixed oxide with alkaline and noble metals created substantial changes of its surface area and pore diameter,with no effect on the crystal dimensions of the main phases or appearance of novel phases in the case of alkaline metals (Table 1).Because the amount of inserted modifiers was relatively small (<5wt %),this could indicate a change of the manganese and cerium oxide crystal packing mode during metal addi-tion.Batch Reactor Experiments.The rate constants and phenol conversions of the different catalysts in batch reactor experiments are presented in Table 2.The manganese oxide component in the form of Mn 3O 4phase yielded a higher catalyst activity relative to that in the form of Mn 5O 8,which was obtained according to a known preparation procedure.22The increase of specific activity by 40%measured for Mn -Ce-1catalyst (vacuum treated)compared with Mn -Ce-2(air treated)corre-sponds to a decrease of the oxidation state of the Mn ions from 3.2(Mn 5O 8)to 2.6(Mn 3O 4).The same trend was observed for Mn -Ce catalysts modified with alka-line metals.To examine the effect of the different phases on the redox cycle,TPO and TPR experiments were performed (Figure 3a,b).According to the TPO spectra (Figure 3a),treatment in a vacuum shifts the oxygen consumption peak to higher temperature as a result ofTable 1.Textural and Structural Properties of Manganese -Ceria-Based Catalystscrystal size (nm)catalyst surface area (m 2‚g -1)pore diameter(nm)pore volume (cm 3‚g -1)phase composition MnO x CeO 2Mn -Ce-19814.80.36Mn 3O 4,CeO 2254Mn -Ce-212811.50.37Mn 5O 8,CeO 2104Mn -Ce -K-17420.00.37Mn 3O 4,CeO 2244Mn -Ce -K-25422.80.25Mn 5O 8,CeO 210 4.5Mn -Ce -Cs-111114.60.40Mn 3O 4,CeO 223 4.5Mn -Ce -Cs-25319.30.25Mn 5O 8,CeO 210 4.5Mn -Ce -Pt 8314.00.32Mn 5O 8,CeO 225 5.0Mn -Ce -Ru7713.30.31Mn 5O 8,CeO 2244.5Figure 2.X-ray diffractograms of Mn -Ce catalysts:(a)Mn -Ce-1,(b)Mn -Ce-2.Table 2.Phenol Conversions and Rate ConstantsMeasured with Mn -Ce-Based Catalysts in Batch Reactor after 60min acatalyst phenol conversion (%)rateconstant,k(10-3L ‚g -1‚min -1)specific rate constant,k ′(10-6L ‚m -2‚min -1)Mn -Ce-1808.990Mn -Ce-2707.765Mn -Ce -K-18713.2178Mn -Ce -K-2728.4155Mn -Ce -Cs-19317.8160Mn -Ce -Cs-2607.8147Mn -Ce -Pt 829.0108Mn -Ce -Ru8310.2130a[C PhO ]0)0.2g ‚L -1,T )100°C,P O 2)10bar,[C cat ])2g ‚L -1.Figure 3.(a)TPO and (b)TPR spectra of Mn -Ce-based catalysts.5092Ind.Eng.Chem.Res.,Vol.43,No.17,2004the lower manganese oxidation state (2.6).In the TPR spectra (Figure 3b),the lack of the low-temperature peak that is assigned to the high-oxidation-state man-ganese species 15corresponded to the formation of the Mn 3O 4phase with a lower oxidation state of Mn ions and a lower dispersion.Most oxygen consumption during O 2treatment and removal during H 2treatment of the evacuated Mn -Ce-1catalyst proceeded at higher temperature (300-350°C)relative to that of air-calcined Mn -Ce-2(150-300°C).This could result in lower efficiency of evacuated Mn -Ce catalyst in phenol con-version to deep oxidation products,facilitating the reactive adsorption of phenol at the catalyst surface.The diminishing redox ability was further observed after addition of alkaline metals.This addition substantially decreased the amount of oxygen adsorbed in TPO experiments by evacuated Mn -Ce catalysts without changing the shape of the TPO (Figure 3a)and TPR (Figure 3b)spectra.Introduction of K and especially of the more basic Cs strongly increased the rate of phenol conversion (Table 2).Treatment in a vacuum along with the addition of alkaline metals increased the rate of phenol oxidative conversion by a factor of 1.6(K)and 2.2(Cs).Introduc-tion of Pt to the more active evacuated Mn -Ce catalyst had no substantial effect on phenol conversion in agreement with the results of Hamoudi et al.13The same result was measured with Ru promoter.Both noble metals increased the specific rate of phenol conversion (normalized per unit of catalyst surface)by a factor of 1.7-2.Contribution of Reactive Adsorption to TOC Removal.The evacuated Mn -Ce-based catalysts that displayed higher activities relative to the air-calcined samples were tested in a fixed-bed reactor.The phenol and TOC conversions and product distributions between solid (due to reactive adsorption)and liquid (due to partial oxidation to low-molecular-weight oxygenates)6,7are shown in Table 3.The parameters were measured in a steady-state parison of the phenol conversions in Tables 2and 3shows that,in both systems (batch and continuous mode),the same ranking of the samples according to their activities was obtained.In all of the continuous-mode experiments,70-94%of the converted phenol was removed through reactive adsorption and CO 2formation.The carbon mass balance based on the amount of carbon detected in the treated water and on that accumulated on the spent catalyst was 96-100%(Table 3).This indicates that negligible CO 2was formed at the selected testing conditions {T )100°C,WHSV (weight hourly space velocity))0.1g of phenol/[(g of catalyst)‚h],LHSV )100h -1},so almost all of the organics removed from the water remained on the catalyst surface,as shown in Table 3.This obser-vation is different from the results reported by Hussain et al.15for phenol CWO in a batch reactor,where the mineralization selectivity toward CO 2formation reached 33%with Mn -Ce and 95%with K -Mn -Ce catalysts even at lower contact times.This difference in product distributions is a result of the different catalyst activa-tion procedure applied,as,in the present study,an evacuation procedure was used instead of air calcina-tion.This yielded a Mn phase with a lower oxidation state of the metal ions and a lower dispersion.Testing of the Mn -Ce-2catalyst that was calcined in air yielded a lower phenol conversion and a reduction of the deposits on the solids surface.In this case,only 68%of the removed organics remained in the catalyst (Table 3).The remaining 32%was converted to CO 2in agree-ment with the data reported previously.13,15Testing of the evacuated Mn -Ce catalysts showed that the alka-line metals increased the phenol conversion and the contribution of reactive adsorption to phenol conversion (Table 3).It was also found that Pt decreased this contribution (in agreement with the results of Hamoudi et al.13),whereas Ru had no effect.Figure 4shows phenol conversions obtained with the evacuated Mn -Ce-1catalyst as a function of phenol load for various LHSVs.Two periods can be identified:(i)a steady-state period at a fixed phenol conversion and (ii)a deactivation period displaying a decline of phenol conversions to e 30%.The deactivation is a result of surface blocking with carbonaceous deposits formed in the catalyst pores by copolymerization of phenol with products of its partial oxidation.7Figure 5and Table 4compare the textural properties of fresh and spent Mn -Ce-1catalysts.After deactivation (phenol conversion e 30%),the wide pore size distribution of the fresh Mn -Ce-1catalyst with a maximum in range of 10-20nm was narrowed,and its maximum was shifted to 4nm because of the formation of a thick carbonaceous depos-its layer on the walls of catalyst pores.This is also confirmed by the decreasing values of the surface area,pore diameter,and pore volume (Table 4).Effects of Additives,LHSV,and Temperature.The steady-state catalyst adsorption capacity (SS-CAC)and the total CAC (measured after X PhOH e 30%)depend on modifiers (K,Cs,Pt,Ru),contact time (LHSV),and temperature,as shown in Tables 5-7.The effect of the addition of alkaline and noble metals on the catalyst operation time,SRA,SS-CAC,and total CAC is shown in Table 5.Addition of basic additivesTable 3.Phenol and TOC Conversions and Product Distributions Measured with Mn -Ce-Based Catalysts before Visible Deactivation aconversion (%)product distribution (%)catalyst phenol TOC liquid solid Mn -Ce-160462377Mn -Ce -K-173621284Mn -Ce -Cs-18275691Mn -Ce -Pt 70493070Mn -Ce -Ru 66522278Mn -Ce-252372950aReaction conditions:[C PhOH ]0)1g ‚L -1,T )100°C,P O 2)10bar,LHSV )100h -1.Figure 4.Phenol conversion as a function of amount of phenol flow through catalyst at different LHSV (T )100°C,P O 2)10bar,[C PhOH ]0)1g ‚L -1,Mn -Ce-1catalyst).Ind.Eng.Chem.Res.,Vol.43,No.17,20045093(K,Cs)reduced both steady-state and deactivation periods,hence reducing catalyst stability.18These find-ings and the steady-state conversions and product distribution presented in Table3indicate that decreased stability is caused by polymer accumulation.The de-crease of the steady-state period and SS-CAC seems to be a result of blocking of the basic sites related to K and especially Cs,being highly reactive in polymeriza-tion reactions.The results for unpromoted Mn-Ce catalyst and that promoted with noble metals,which are shown in Table5,were similar for the steady-state region(duration of operation period and SS-CAC). However,both the deactivation period and the total CAC were smaller with the promoters.This is in agreement with the results of Hamoudi et al.,13who found that Pt-promoted Mn-Ce catalyst is sensitive to deactivation.The effect of increasing LHSV on phenol and TOC conversions,CAC,and SRA are presented in Table6 for the Mn-Ce-1catalyst.The increase in LHSV had little effect on SS-CAC and total CAC.It substantially affected the period of steady-state operation,phenol conversion,and SRA,as shown in Table6.In addition, as the space time increased,the phenol transformation gradually shifted to reactive adsorption.The increase of space time(lower LHSV)also decreased the specific load(grams of phenol on stream per gram of catalyst) as shown in Figure4,allowing longer operation time and higher phenol conversion.On the basis of the TOC balance,it can be concluded that the higher phenol conversion caused by higher amounts of carbonaceous solids deposited on the catalyst surface(Table6)also results in a higher SRA.It is important to mention that the Mn-Ce-1catalyst at100°C removed95-100% phenol from contaminated water at LHSVs of10-1h-1 with SS-CACs of120-130mg of phenol/g of catalyst (Table6).The effect of temperature on the phenol and TOC conversions and on the CAC and SRA are presented in Table7for the Mn-Ce-1catalyst.Decreasing the reaction temperature from130to80°C for a steady LHSV gradually enhanced the SRA,SS-CAC,and total CAC.The same trend was obtained when the LHSV was decreased for a steady temperature.Although the phenol conversion decreased with temperature,the selectivity toward reactive adsorption(SRA)increased, due to the fact that there is a higher amount of unreacted phenol that is free to be adsorbed on the catalyst surface(which has a high affinity to adsorb carbonaceous deposits).In this process,more polymeric deposits are adsorbed with lower deactivation ability yielding an increase of SS-CAC and total CAC to170 and420mg of phenol/g of catalyst,respectively(Table 7).Decreasing SS-CAC,total CAC,and SRA with increasing temperature,in the absence of CO2emission, corresponded to production of more liquid products and not to oxidation of carbonaceous deposits.For a deeper understanding of the observed effects of additives,LHSV,and temperature on the performance of the Mn-Ce catalysts and the reactive-adsorption phenomena,it will be helpful to collect direct informa-tion about the nature of carbonaceous deposits as aFigure5.Pore size distributions of fresh,spent,and regenerated Mn-Ce-1catalyst.Table4.Textural Properties of Mn-Ce-1Catalystsample surface area(m2‚g-1)pore diameter(nm)pore volume(mL‚g-1)fresh9814.80.36 spent56 3.80.05 regenerated11511.20.30 Table5.Steady-State and Deactivation Periods,Phenol Adsorption Capacities,and Selectivities toward Reactive Adsorption as a Function of Catalyst Composition acatalyst steady-stateperiod(h)deacti-vationperiod(h)SS-CAC(mg ofphenol/g ofcatalyst)total CAC(mg ofphenol/g ofcatalyst)SRA(mg ofphenol/g of reactedphenol)Mn-Ce-1 2.514.4120300500Mn-Ce-K-10.610.540310670Mn-Ce-Cs-10.6550320810Mn-Ce-Pt 2.56130270550Mn-Ce-Ru 2.5 4.2150240660a Reaction conditions:LHSV)100h-1,[C PhO]0)1g L-1,T) 100°C,P O2)10bar.Table6.Phenol Adsorption Capacity and Reactive Adsorption Selectivity of Mn-Ce-1Catalyst as aFunction of Catalyst Composition aLHSV(h-1):11050100 phenol conversion b(%)100957560 TOC conversion b(%)100886246 SS-CAC(mg of phenol/g of catalyst)120130130120 total CAC(mg of phenol/g of catalyst)300300290300 SRA(mg of phenol/g of reacted phenol)1000690500500 steady-state period(h)7876 2.5 deactivation period(h)340402514.4 product distribution b(%)liquid071723 solid100938377a Reaction conditions:[C PhO]0)1g‚L-1,T)100°C,P O2)10bar.b Before visible deactivation.Table7.Phenol Adsorption Capacity and Reactive Adsorption Selectivity of Mn-Ce-1Catalyst as aFunction of Reaction Temperature atemperature(°C):LHSV(h-1):80101001080100100100130100 phenol conversion b(%)8995526075 TOC conversion b(%)8488474647 SS-CAC(mg of phenol/g of catalyst)160130170120110 total CAC(mg of phenol/g of catalyst)420300320300280 SRA(mg of phenol/g of reacted phenol)790690530500410 product distribution b(%)liquid57102337 solid9593907763a Reaction conditions:[C PhO]0)1g‚L-1,P O2)10bar.b Before visible deactivation.5094Ind.Eng.Chem.Res.,Vol.43,No.17,2004。