Practical synthesis of aromatic amines by photocatalytic reduction of aromatic nitro compounds on nanoparticles N-doped TiO 2Huqun Wang a ,Junping Yan a ,Wenfu Chang b ,Zhimin Zhang a,*a School of Chemistry and Chemical Engineering,Shanxi University,Wucheng Road,Taiyuan 030006,PR China bInstitute of Molecular Science,Shanxi University Taiyuan 030006,PR Chinaa r t i c l e i n f o Article history:Received 9September 2008Received in revised form 15December 2008Accepted 17December 2008Available online 25December 2008Keywords:ReductionAromatic amines N-doped TiO 2Potassium iodidea b s t r a c tA novel efficient method for the catalytic reduction of aromatic nitro compounds to the corresponding amines was reported.Aromatic nitro compounds were chemoselectively reduced to the corresponding amines by using N-doped TiO 2and potassium iodide as photocatalysts in the presence of methanol.The novel method is highly efficient with very short reaction time (<20min),excellent yields (>90%)and wide functional group tolerance such as carbonyl,halogen,amino,hydroxyl and carboxylic acid groups.And N-doped TiO 2was prepared by a modified sol-gel method using urea as nitrogen source and had higher photocatalytic activity comparing with pure TiO 2.The catalysts were characterized by XRD,XPS,TEM and UV–Vis DRS.Ó2009Published by Elsevier B.V.1.IntroductionAromatic amines are widely used key intermediates in the industrial synthesis of dyes,pharmaceuticals and agrochemicals [1].A variety of methods for the direct reduction of aromatic nitro compounds to the corresponding amines has been well docu-mented [2–4].However,development of new methodology espe-cially the environmentally benign process still attracts the great interests in the chemistry community [5–8].In comparison to the commonly used methods which involve hydrogenation,elec-tron transfer and hydride reduction,photocatalytic reduction emerges as cost-effective,highly selective,rapid and environmen-tally friendly.Li and co-workers first reported a photoinduced reduction of nitro compounds to the corresponding amines using TiO 2semiconductor as a catalyst [9].Ferry and co-workers further investigated the mechanism of photocatalytic reduction of nitro aromatics at the surface of titanium dioxide slurries in the pres-ence of the sacrificial electron donor methanol or isopropanol [10].Heterogeneous photocatalysis has been rapidly becoming an exciting and growing area of research due to its direct application for synthetic chemistry,such as Photo–Kolbe oxidation [11],reduc-tion [12],amino acid [13],Diels–Alder [14]and Friedel–Crafts alkylation [15]reactions.However,so far these reactions are still hardly applied to the industrial field.Research in our laboratory has focused in the late few years on new active-TiO 2based reduc-ing systems.The reaction for synthesis of amines from nitro com-pounds was catalyzed by irradiating N-doped TiO 2(N-TiO 2)and potassium iodide in solution of methanol (Scheme 1).The proce-dure of synthesis of aromatic amines was much simpler and more efficient than those in any other literature.In addition,the photo-catalyst could be reused and remained sufficient catalytic activity.2.Experimental2.1.Catalyst preparation and characterizationAll reagents were analytical reagent grade and were used with-out any further purification.A solution of tetrabutyltitanate (8.5mL)in absolute ethanol (30mL)was mixed with glacial acetic acid (1.5mL)as constraining reagent to prevent the precipitation of oxides and stabilize the solution and an ethanol solution of dis-tilled water and urea (EtOH:H 2O:CH 4ON 2=3:48:1)was added to above solution under vigorous stirring.The pH of solution was ad-justed about three by nitric acid.After 3h,the gel so obtained had been left ageing overnight at room temperature to ensure the com-pletion of the hydrolysis,subsequently evaporation of the solvent,drying at 100°C for 8h and finally calcination at 450°C for 4h.The anatase crystal phase was determined from the X-ray diffraction (XRD)patterns obtained by using an X-ray diffractometer (Model D/Max 2550V)with a Cu target Ka-ray (k =1.544178Å).The mor-phology of the N-TiO 2powders was examined by using a Hitachi-600-2transmission electron microscope (TEM)and UV–Vis diffuse reflectance spectrophotometer (Cary 300,Varian,US)was employed to determine the optical properties of N-TiO 2and pure1566-7367/$-see front matter Ó2009Published by Elsevier B.V.doi:10.1016/j.catcom.2008.12.045*Corresponding author.Tel.:+863517010588;fax:+863517011688.E-mail address:mqz1003@ (Z.Zhang).Catalysis Communications 10(2009)989–994Contents lists available at ScienceDirectCatalysis Communicationsjournal homepage:www.elsev i e r.c o m /l o c a t e /c a t c omTiO 2.The surface composition and bonding configuration were measured by XPS in a VG ESCALAB250system.2.2.Photocatalytic synthesis of aromatic aminesPhotocatalytic reactions were carried out in a cylindrical round bottomed quartz photo-reactor and irradiated using 500W high-pressure mercury lamp and solar light under magnetic stirring at room temperature.The reduction of aromatic nitro compounds (1mmol)was carried out in the present of N-TiO 2(10g/L)and potassium iodide (0.3mmol)in methanol (30mL),which was irra-diated for 6–20min.Then while stirring,the reaction mixture be-came heterogeneous as the reaction progressed.The light yellow N-TiO 2was filtered and the filtrate was concentrated to dryness.The crude product was further purified by silica-gel column chro-matography to give the product amines.All products gave HPLC spectra consistent with the spectra of standard.3.Results and discussion3.1.TEM observations and optical properties of N-TiO 2powders Fig.1a shows typical TEM images of anatase N-TiO 2particles.The characterization of catalyst by TEM studies showed that the nanoparticles were well-distributed.The average size of nitrogen doped TiO 2nanoparticles was between 5.20and 13.5nm which were essentially consistent with the XRD result.The size distribu-tion graphic was depicted in Fig.1b.The UV–Vis DRS absorption spectra of N-TiO 2and TiO 2were shown in Fig.2.N-TiO 2sample absorbed more visible light in the range of 400–550nm comparing with pure TiO 2,and the method was effective to shift the optical absorption of N-TiO 2particles to-ward the visible regions.The band gap of sample is determined by the equation E g =1239.8/k [16],where E g (eV)is band gap and k (nm)is the wavelength of the absorption edge in the spectrum.The band gaps obtained optically were approximately 2.85and3.02eV for N-doped TiO 2and TiO 2,respectively,revealing that the band gap of TiO 2was narrowed by N doping.The absorption edge of N-TiO 2was moved to 550nm,resulting in extending the activating spectrum from UV to the visible range.The absorption edge was shifted to the lower-energy region in the spectrum of N-TiO 2.The band gap narrowing in our research may be caused by introduction of nitrogen from urea into the lattice of titanium.Thus,the sample of N-TiO 2showed the excellent visible light absorption,indicating the increase of photocatalytic activity in vis-ible region.3.2.XRD observations and XPS analysisShown in Fig.3were X-ray powder diffraction patterns of TiO 2and N-TiO 2powder samples prepared.From the intensity ratios be-tween the diffraction appearing at 2h =25.5(anatase 101)may conclude that the TiO 2deposited at 450°C consists almost com-pletely of the anatase phase and no other crystal phase can be de-tected.The average crystalline size of N-TiO 2was calculated using the Scherrer equation.Pure TiO 2had a particle size of 10.4nm;doped TiO 2,about 6.80nm.N-doped TiO 2powers showed smaller size than undoped-TiO 2prepared at the same calcinationtemperature.Fig.1.(a)TEM micrograph of N-TiO 2nanoparticles.(b)Size distribution of N-TiO 2nanoparticles determined by TEM.The sizes were determined for 100nanoparticles selected at random.990H.Wang et al./Catalysis Communications 10(2009)989–994Shown in Fig.4were XPS spectra of the as-synthesized N-TiO2 samples.The N1s and Ti2p were found from the XPS patterns of N-TiO2.It could be seen from Fig.4a that the broad N1s peak was found at399.8eV,showing that nitrogen was incorporated into the TiO2.The N feature at a binding energy of396.6eV was mostly assigned as the substitutional incorporation of nitrogen into the TiO2lattice,and the peak at399.8eV was assigned as N A O and N A N bonding,or the nitrogen chemically bound to hydrogen when introduced by N dopants.Chen and Burda[17]ob-served a broad XPS peak centered at401.3eV for their N-doped TiO2nanoparticles and they attributed it to the substitutional O–Ti–N sites in the TiO2lattice.However,the nitrogen state in the doped TiO2may vary form case to case and variations in the XPS results may be associated with the different surface structures. So we could conclude the Ti–O–N bonds had been formed.In the Ti2p region,both the Ti2p2/3and Ti2p1/2were found at461.9 and467.6eV,respectively.According to the XPS standard spec-trum[18],the Ti2p2/3peak of TiO2should be at458.8eV,but from Fig.4b the Ti2p2/3was found at461.9eV,and so,it was confirmed that Ti in N-TiO2did not exist in the form of TiO2only.Miao et al.[19]also assigned these shifted peaks to TiO2Àx N x.Based on the above XPS analysis,the broad N1s feature and the shift of Ti2p peaks in this work provide strong evidence for the substitutional incorporation of nitrogen into the TiO2and the TiO2Àx N x structure was well formed.3.3.Photocatalytic activityThe photocatalytic activity of doped and undoped nanoparticles was investigated by the photocatalytic reduction of o-nitrophenol as a model of reaction.The P25,pure TiO2and N-TiO2were irradi-ated by UV light.Results were shown in Fig.5.Under UV irradia-tion,the yield of photocatalytic synthesis of o-aminophenol was about10%with P25.This was anticipated because the activity of P25had been proved to be very low.However,the yield with N-TiO2was very high(99.27%)after reaction for7.5min,compar-ing with pure TiO2(40.05%).Similarly,we also investigated the photocatalytic activity under solar light and the results were shown in Fig.5.As shown in Fig.5,no reducing products were ob-tained with P25under solar light after300min,while the yield was also very small(3.24%)with pure TiO2in the same time. According to the previous report,pure TiO2had no photocatalytic activity under visible light.So this may be attributed to the pres-ence of ultraviolet in solar light.Then we made the supplement experiment under solar light through UV-cutfilter sheet and the result showed no products were detected on pure TiO2after 300min.The N-TiO2still showed the best catalytic activity among the catalysts under solar light.As a comparison,we also measured the photocatalytic performance of potassium iodide under the same experimental conditions.Nevertheless,no photo-reduction products were obtained with KI alone under UV irradiation.Fur-ther experiments were carried out to confirm whether the reaction was photocatalytic or not and it was observed that no products were formed in the absence of N-TiO2catalyst.It was therefore be-lieved that the interactions between N-TiO2and KI contributed toH.Wang et al./Catalysis Communications10(2009)989–994991992H.Wang et al./Catalysis Communications10(2009)989–994 Table1the superior activity for the reduction of nitroarenes.Based on the experiments above,it was obvious that N-TiO2had higher photo-catalytic activity comparing with pure TiO2and P25,which was as-cribed to the N dopant.This was consistent with the analytical results.According to the XRD analysis,the N dopant of TiO2de-creased the crystalline size of TiO2which resulted in the enhance-ment of photocatalytic activity.According to XPS analysis,the entry of N into the TiO2lattices suppressed the particle growth and consequently caused a decrease of oxygen vacancies,whichminimized the electron-hole recombination during the photocata-lytic synthesis of aromatic amines.3.4.Photocatalytic synthesis of aromatic aminesThe reduction for aromatic nitro compounds has been carried out using the procedure under UV irradiation because of the long reaction time under solar light.Quantitative analysis of the yield for the reaction was performed in triplicate and the standard devi-ation was calculated.The results were listed in Table1.The data presented clearly showed the efficiency of potassium iodide and N-TiO2in the reduction of several nitro compounds.Moreover, the reaction time was much shorter than those reported in any other literature.It could be seen that the reduction of nitroarenes gave corre-sponding amines in excellent yield(above90%).The catalytic sys-tem was efficient in the reduction of aromatic nitro compounds bearing additional substituents in aromatic ring.With regard to the reactivity of the different nitroarenes tested,it was worth not-ing that the electronic properties of substituents attached to the aromatic ring have any effect on the reduction of nitro groups. Moreover,nitro compounds bearing electron-withdrawing groups, which have been reported as more reactive or active when reacted, gave the corresponding amines in better yields and in shorter reac-tion time comparing with that of the aromatic nitro compounds substituted with strong-electron releasing groups.Concerning the steric hindrance effect of nitro groups with substituent at their ortho position,we have observed no adverse affect for the reduc-tive process in the case of ortho substituent that were expected less reactivity such as hydroxyl,methyl and amino at this position and this was inconsistent with the conventional principle.How-ever,as shown in Table1,the meta position required a longer irra-diation time compared with the ortho position and this was also different from what would be expected.The effect of the additives by reduction of o-nitrophenol as a model reaction has been checked out by comparing preliminary yields with additives and without additives.Potassium iodide and sodium iodide were found to be the most effective additives (Table2).Indeed,using an additive could improve the yield of reac-tion.The probable reaction mechanism was shown in Scheme2. With UV or high energy visible light,an excited state iodide ion would be probably produced.This together with the solvent led to the electron cage complex which,in turn,reacts either with an oxidizing agent or a proton donor.In the process of reaction,meth-anol provided the proton and produced the hydrogen radicals, while TiO2produced pairs of electron-holes(eÀ,h+)under irradia-tion,and the iodide anion was oxidized into the iodine by parts of holes.Accordingly,iodide ion was a scavenger that reacts with po-sitive holes,decreasing the number of oxidizing species available on the catalyst surface.However,with the increase of the KI con-centration,iodide anion competed for the adsorptive sites on the N-TiO2surface with o-nitrophenol,resulting in the decrease of yield.With KI alone the reduction reaction didn’t occur and it may be interpreted that the iodide ion was not oxidized into the molecular iodine without conventional oxidant,because the ex-cited state iodide ion would be prone to return the base state.That is,it will be not completely effectively carried out for the reduction of molecular iodine by organic matter,and hence molecular iodine should not appear.4.ConclusionsIn conclusion,we have described herein a new methodology which can efficiently convert aromatic nitro compounds in metha-nol using N-TiO2and potassium iodide,under more facile and inex-pensive conditions than those described in a previous work.The reagent system described here has been proved to be a good alter-native to well known methods of reduction of aromatic nitro com-pounds.Most interestingly the N-TiO2photocatalyst is stable and reusable under photo-irradiation in the repeated experiments.No meaningful difference in reaction yields is observed in the three re-peated photo-reduction experiments.New applications of this reducing process are being developed.HPLC yield.b SD,standard deviation.Table2Reduction of o-nitrophenol with N-TiO2in the presence of additives(n=3).Entry Additive Reaction time(min)Yield a(mean±SD b)(%)17Without additive1095.47±0.3518KI7.599.27±0.2119KBr1194.90±0.2620NaI7.599.30±0.3621NaBr1093.70±0.8222Na2SO41095.07±0.1223C6H7SO3Na1294.73±0.25a HPLC yield.b SD,standard deviation.H.Wang et al./Catalysis Communications10(2009)989–994993AcknowledgementsWe thank the Science and Technology of Shanxi Province,China (Project No.2006031141)forfinancial assistance and nature sci-ence fund of Shanxi University.References[1]J.S.D.Kumar,M.H.ManKit,T.Toyokuni,Tetrahedron Lett.42(2001)5601.[2]C.Kazmierski,J.M.P.Gosmini,J.Périchon,Tetrahedron Lett.44(2003)6417.[3]Y.Moglie,C.Vitale,G.Radivoy,Tetrahedron Lett.49(2008)1828.[4]U.Stéphane,A.Falguières,A.Guy,C.Ferroud,Tetrahedron Lett.46(2005)5913–5917.[5]A.R.Gandhe,J.B.Fernandes,Bull.Catal.Soc.India4(2005)31.[6]T.C.Jagadale,S.P.Takale,R.S.Sonawane,et al.,J.Phys.Chem.C112(2008)14595.[7]M.Sathish,B.Viswanathan,R.P.Viswanath,Appl.Catal.B:Environ.74(2007)307.[8]K.Abiraj,G.R.Srinivasa,D.C.Gowda,Aust.J.Chem.58(2005)149.[9]F.Mahdavi,T.C.Bruton,Y.Z.Li,.Chem.58(1993)744.[10]L.John,H.G.FerryWilliam,Langmuir14(1998)3551.[11]B.Kraeutler,A.J.Bard,J.Am.Chem.Soc.100(1978)2239.[12](a)L.Lin,R.R.Kuntz,Langmuir8(1992)870.[13](a)W.W.Dunn,Y.Aikawa, A.J.Bard,J.Am.Chem.Soc.103(1981)6893.[14]W.Zhang,X.D.Jia,L.Yang,Z.L.Liu,Tetrahedron Lett.43(2002)9433.[15]M.L.Kantam,ha,J.Yadav, B.Sreedhar,Tetrahedron Lett.47(2006)6213.[16]K.S.Rane,R.Mhalsiker,S.Yin,T.Sato,K.Cho,E.Dunbar,P.Biswas,J.Solid StateChem.179(2006)3033.[17]X.Chen,C.Burda,J.Phys.Chem.B108(2004)15446.[18]Y.Liu,D.Z.Sun,Appl.Catal.B:Environ.72(2007)205.[19]L.Miao,S.Tanemura,H.Watanabe,Y.Mori,K.Kaneko,S.Toh,J.Cryst.Growth260(2004)118.994H.Wang et al./Catalysis Communications10(2009)989–994。