Chemical Engineering Journal

Transcription

1 Chemical Engineering Journal 172 (211) Contents lists available at ScienceDirect Chemical Engineering Journal j ourna l ho mepage: Selective oxidation of benzyl alcohol to benzaldehyde in water by TiO 2 /Cu(II)/UV solar system Raffaele Marotta, Ilaria Di Somma, Danilo Spasiano, Roberto Andreozzi, Vincenzo Caprio Department of Chemical Engineering, Faculty of Engineering, University of Naples Federico II, p.le V. Tecchio, Naples, Italy a r t i c l e i n f o Article history: Received 11 February 211 Received in revised form 25 May 211 Accepted 25 May 211 Keywords: Benzylic alcohol Selective photocatalytic oxidation Benzaldehyde production Titanium dioxide Cu(II) reduction a b s t r a c t Selective oxidation of benzyl alcohol to benzaldehyde in aqueous solution, under acidic conditions, through the TiO 2 /Cu(II)/solar UV photocatalytic system, was investigated. Different commercial TiO 2 samples were tested. The best result found, in terms of yield, was of 35% for benzaldehyde with respect to the initial benzyl alcohol concentration. During a single run, a partial conversion of benzaldehyde to benzoic acid was also observed. By-products, present at trace level, were 2-hydroxy-benzyl-alcohol, 4-hydroxy-benzylalcohol, 2-hydroxy-benzaldehyde and 4-hydroxy-benzaldehyde. On the basis of the formation of these species a production of HO radicals could be thus inferred. The study suggested that different operative parameters, such as the composition and amount of photocatalyst, ph, ionic components in water and the initial concentration of Cu(II) ions, played an important role in the photocatalytic selective oxidation of benzyl alcohol. Moreover, the results of the present investigation indicated that, at the end of the process, Cu(II) could be regenerated and reused, through a re-oxidation of Cu(), produced during the photolytic run, with air in the dark. A general mechanism of oxidation, supported by the experimental results was proposed. 211 Elsevier B.V. All rights reserved. 1. Introduction Titanium oxide (TiO 2 ) photocatalysis is one of the most studied processes to oxidize organic molecules in aqueous solution by means of a solar UV source and oxygen (or air) [1 3]. It has been extensively investigated the reaction mechanism that makes the process occurr. It includes, as the key step, the photochemical formation of an electron hole pair on TiO 2 particles which starts the process itself [4]. Titanium oxide photocatalysts have a great potential also as a versatile tool in green organic synthesis [5]. In recent years, TiO 2 photocatalysis has been successfully proposed replacing water with CH 3 CN as a selective oxidation process of aromatic alcohols to aldehydes, with yields approaching to 1%, relying on the refractoriness of the latter to the oxidation by positive holes [6,7]. However, since water remains an elective solvent for environmentally friendly processes, other approaches are necessary in order to avoid the use of organic solvents. A possibility could rely on the use of Cu(II) ions as electron acceptors to replace oxygen in a TiO 2 photocatalytic process in water. The idea supporting this choice is that, in the absence of oxygen, it is inhibited one of the photochemical pathways leading to the unselective OH radical pro- Corresponding author. Tel.: ; fax: address: rmarotta@unina.it (R. Marotta). duction [8,9]. In fact, in the absence of oxygen, it is not possible the formation of superoxide radical (O 2 ) or hydrogen peroxide (H 2 O 2 ) whose photolysis can generate OH radicals; therefore, in this case, OH radicals should form just as result of the reaction of water with positive holes. It is interesting to note that a different approach, based on the use of nanostructured TiO 2 catalysts, leads to very interesting results although, at the moment, an explanation of them is well beyond the ability of the Authors [1,11]. It is well known that the substitution of oxygen with a species capable of reducing, by trapping the electrons in the conducting band, still enables the oxidation of the organic species present in the solution. Particularly interesting is the case in which oxygen is replaced by a metal ion dissolved in the solution. The latter reduces to a lower oxidation state by capturing the photo-generated electrons on TiO 2, whereas the organic species oxidizes, through a direct reaction with the positive holes (h + ) or OH radicals. In some cases, the reduction of the metal results into its precipitation from the solution thus enabling its separation and recovery. If this is the proper aim of the process, the role of the organic species is that of a sacrificial agent whose oxidation makes possible the reduction and the recovery of the metal. The reduction of many metals, such as Cu(II), Ni(II), Pb(II) and Zn(II), has been investigated in the past by some researchers [12,13]. Among the others, some papers have been devoted to the photoreduction of Cu(II) in the presence of organic species such as formic or oxalic acids [14,15], methanol, /$ see front matter 211 Elsevier B.V. All rights reserved. doi:1.116/j.cej

2 244 R. Marotta et al. / Chemical Engineering Journal 172 (211) Scheme 1. Mechanism depicting the simultaneous photocatalytic oxidation of organic species and the reduction of Cu (II). n-propanol, 2-propanol and n-butyl alcohol [16] in absence of oxygen in the system. It has been reported that the simple admission of air into the system, at the end of the process, allows the re-oxidation of precipitated copper which is, in this way, completely solubilised [15]. The nature of the precipitate has never been completely elucidated, Cu() and/or Cu 2 O being reported by the most part of researchers [12,15 19], although some of the Authors reported some results in favour of the formation of Cu() [14]. The overall behaviour of the system, including the reoxidation of copper, is illustrated in Scheme 1. A two stages oxidation process results from the scheme: in the first stage, the reduction of copper(ii) allows the oxidation of the organic compound, in the second one the air reoxidizes the reduced form of copper. It is evident from this scheme that once the reoxidation of reduced copper to Cu(II) by air is taken into account, the role of copper (in the presence of TiO 2 and UV) is that of a catalytitc species (moving between two (or three) different oxidation states ( or +1 and +2), which makes possible the oxidation of the organic substances by air at ambient conditions. On the basis of this consideration, the use of Cu(II)/TiO 2 /UV as a photocatalytic system to oxidize organic species by air can be proposed. At the best of the Authors knowledge no previous investigations have been reported on the use of this system for selective oxidation of organic molecules. Benzaldehyde is the simplest and industrially the most important aromatic aldehyde. Benzaldehyde is used in a large number of applications; among the others, as an industrial solvent and commercial food flavouring and it is an interesting key intermediate for various perfumes and dyes. Benzaldehyde is produced principally by the hydrolysis of benzylidene chloride or the partial oxidation of toluene [2]. Both these processes employ very hard operating conditions and cause wastewater disposal problems. Fattima Al-Zahra Gassima and coworkers [21] investigated the possibility to convert liquid benzyl alcohol to benzaldehyde through a suspension of titanium dioxide (anatase) and sensitized anatase under an oxygen atmosphere but using p-xylene as solvent for benzyl alcohol. In the present work the possibility to employ the Cu(II)/TiO 2 /UV system for the selective oxidation of benzyl alcohol to benzaldehyde in water is studied at varying the operating conditions (TiO 2 type, TiO 2 load, nature of the inorganic anions, Cu(II) concentration and ph). 2. Materials and methods Experiments have been carried out in a batch cylindrical glass jacketed reactor (28 ml) equipped with high-pressure UV lamp (Helios Italquartz) mainly emitting at 5, 313 and 366 nm. The reactor has been thermostated at 298 K. The ph has been regulated with phosphoric acid and monitored by means of an Orion 42A+ ph-meter (Thermo). In all the exper Fig. 1. Photooxydation of benzyl alcohol effect of TiO 2 type. [Cu(II)] o = 1.5 mm. [Benzyl alcohol] o = 1.5 mm. ph = 2.. T = 25 C. [TiO 2] o = 2 mg/l. Runs without oxygen: Aldrich (pure anatase, SA = 9.5 m 2 g 1 ), P25 Degussa (8% anatase, SA = 5 m 2 g 1 ), Aldrich (pure rutile, SA = 2.5 m 2 g 1 ), Aldrich (prevalently rutile, SA = 2.7 m 2 g 1 ). Run with oxygen without Cu(II): Aldrich (pure anatase). iments the solution has been preventively purged with nitrogen to remove the dissolved oxygen that could have competed with cupric ions for the reaction with the electrons. During the runs a gaseous stream of nitrogen has been continuously fed to the irradiated magnetically stirred solution to prevent any contact with oxygen. The concentrations of benzyl alcohol, benzaldehyde, benzoic acid, 2-hydroxy-benzyl alcohol, 4-hydroxy-benzyl alcohol, 2- hydroxy-benzaldehyde and 4-hydroxy-benzaldehyde at different reaction times have been evaluated by HPLC analysis. For this purpose, the HPLC apparatus (Agilent 11) has been equipped with a diode array UV/Vis detector ( = 22, 2, 25 nm) and a sinergy 4 Hydro-RP 8A (Phenomenex) column, using a mobile phase of 5% buffer, % H 2 O and 2% CH 3 CN, flowing at 1. ml min 1. One litre of buffer has been made by 1 ml of phosphoric acid solution (5.5 M), 5 ml of methyl alcohol and water for HPLC. The concentration of cupric ions has been measured by means of a colorimetric method using an analytical kit (based on oxalic acid bis-cyclohexylidene hydrazide, cuprizone ) purchased from Macherey-Nagel. An UV/Vis spectrometer (Unicam) has been used for the measurements at a wavelength of 585 nm. Four commercial microcrystalline TiO 2 powders have been studied: (1) TiO 2 Degussa P25 (8% anatase, 2% rutile, BET specific surface area 5 m 2 g 1 ), (2) TiO 2 Aldrich (pure anatase phase, BET specific surface area 9.5 m 2 g 1 ), (3) TiO 2 Aldrich (pure rutile phase, BET specific surface area 2.5 m 2 g 1 ), (4) TiO 2 Aldrich (rutile phase with small amount of anatase, BET specific surface area 2.7 m 2 g 1 ). BET specific surface areas have been measured by the singlepoint BET method using a Micrometrics Flow Sorb 22 apparatus. Cu(II) ions have been introduced in the system as cupric sulphate. Benzyl alcohol, benzoic acid, phosphoric acid, cupric sulphate, sodium sulphate and sodium dihydrogen phosphate, with a purity >99.% (w/w), have been purchased from Sigma Aldrich as well as benzaldehyde with a purity >9% (w/w) and used as received

3 R. Marotta et al. / Chemical Engineering Journal 172 (211) Fig. 2. Effect of TiO 2 type: benzaldehyde (a) and benzoic acid (b) production. [Benzyl alcohol] o = 1.5 mm. [Cu(II)] o = 1.5 mm. ph = 2.. T = 25 C. [TiO 2] o = 2 mg/l. Only stripping with nitrogen at dark ( ). Runs without oxygen: Aldrich (pure anatase, SA = 9.5 m 2 g 1 ), P25 Degussa (8% anatase, SA = 5 m 2 g 1 ), Aldrich (pure rutile, SA = 2.5 m 2 g 1 ), Aldrich (prevalently rutile, SA = 2.7 m 2 g 1 ). Run in presence of oxygen without Cu(II): Aldrich (pure anatase). 3. Results and discussion Preliminary photolytic runs (data not shown) in presence of the substrate and Cu(II) ions without TiO 2 or the substrate and TiO 2 without Cu(II) addition to the solution under nitrogen gaseous stream, did not result into any consumption of benzyl alcohol even for long reaction times (more than 12 h) Effect of TiO 2 type The results obtained during some runs of photoxidation of benzyl alcohol at ph = 2., with different TiO 2 commercial samples, at a load equal to 2 mg/l, are shown in Fig. 1. The diagrams show that the reactivity of benzyl alcohol is strongly influenced by the type of TiO 2 used in the experiment, whereas the specific surface area could not be taken as a predictor of the sample reactivity. In particular, the best results in terms of conversion, at a fixed reacting time, are observed when Aldrich TiO 2 (pure anatase) is used in the run. In this case, after 12 min of reaction, the concentration of Cu(II) approaches to zero with a conversion of the alcohol of about 75%, despite of an initial ratio [benzyl alcohol/cu(ii)] = 1. A similar reactivity, but with a lower conversion of benzyl alcohol (65%), is shown by a P25 Degussa sample in which anatase form is present at 8% with a specific surface area (5 m 2 g 1 ) higher than the Aldrich sample (pure anatase, 9.5 m 2 g 1 ). The catalysts in which TiO 2 is present as rutile form, either prevalently or totally, show lower activities than the samples containing prevalently anatase. The observed results can be explained through the fact that for both crystallographic forms of TiO 2 (anatase and rutile), the valence band (VB) redox potentials are more positive (2.96 and 2.85 V vs NHE respectively) [27] than the OH/ OH and OH/H 2 O redox couples (1.89 and 2.72 V vs NHE respectively) [28]. Consequently, both adsorbed water and hydroxyl groups can be oxidized to reactive hydroxyl radicals on both irradiated TiO 2 types surfaces [1]: Ti IV OH + h + Ti IV OH (r1) Ti IV H2O + h + Ti IV OH + H + (r2) Nevertheless the more negative redox potential (.27 V vs NHE) of the anatase conduction band (CB), makes it more competitive than the rutile one (.15 V vs NHE) for reduction reactions [27]. In this sense, taking into account that the standard redox potential of Cu 2+ /Cu() couple is.337 V [9], cupric ions can be reduced to metal copper by anatase CB electrons more easily than by rutile CB electrons. In any case, as reported by others [22,23], several other properties of the tested photocatalysts such as particle geometry, crystallinity, density of surface functional groups, and defects, should be considered to foresee the behaviour of the adopted TiO 2. Moreover, since for each Cu(II) ion reduced to Cu(), two photogenerated electrons are consumed and two positive holes have to be saturated, the experimental data invariably indicate the existence of secondary reactions in which other species, in competition with benzyl alcohol molecules, consume the positive holes thus reducing the consumption ratio [benzyl alcohol]/[cu(ii)], to a value lower than 1.. As it can be inferred from Fig. 2a, during the process the substrate is mainly converted into benzaldehyde that partially undergoes to a further oxidation to benzoic acid (Fig. 2b). When the highest conversion of the substrate is achieved (72%), 35% of initial benzyl alcohol resulted to be converted into benzaldehyde and only 8% into benzoic acid for pure anatase TiO 2. In order to better understand the importance of replacing oxygen with Cu(II) ions, the data related to a run in which the alcohol is contacted with oxygen, without any addition of Cu(II), in the presence of Aldrich TiO 2 (pure anatase) sample are presented in Figs. 1 and 2a (empty circles). Although the system results to be capable of promptly converting benzyl alcohol (Fig. 1), the yield in benzaldehyde is very low (Fig. 2a). Similar results have been also obtained with different TiO 2 samples (data not shown). The slight decrease of the concentration profiles of benzaldehyde for reaction times higher than 9 min (Fig. 2a: full diamonds, full circles), when Cu(II) is totally converted to Cu(), could be ascribed to its loss from the solution due to a stripping effect of the inlet nitrogen gaseous stream bubbling. An example of the importance of this effect is given in the same figure (Fig. 2a: empty squares) reporting the results collected by bubbling a nitrogen

4 246 R. Marotta et al. / Chemical Engineering Journal 172 (211) [Cu(II)]/[Cu(II)]o.6 Normalized concentration.6 Nitrogen bubbling Light on Oxygen bubbling Light off Nitrogen bubbling Light on Fig. 3. Photoreduction of copper(ii) effect of TiO 2 type. [Cu(II)] o = 1.5 mm. [Benzyl alcohol]o = 1.5 mm. ph = 2.. T = 25 C. [TiO 2] o = 2 mg/l. Runs without oxygen: Aldrich (pure anatase, SA = 9.5 m 2 g 1 ), P25 Degussa (8% anatase, SA = 5 m 2 g 1 ), Aldrich (pure rutile, SA = 2.5 m 2 g 1 ), Aldrich (prevalently rutile, SA = 2.7 m 2 g 1 ). gaseous stream in a solution initially containing only benzaldehyde (at dark and in absence of TiO 2 and Cu(II) ions). Small amounts of undesired by-products, such as 2-hydroxy-benzyl-alcohol, 4-hydroxy-benzyl-alcohol, 2-hydroxybenzaldehyde and 4-hydroxy-benzaldehyde, have been detected during the runs in the reacting solutions, as a result of hydroxyl radicals (HO. ) attack to benzyl alcohol and benzaldehyde molecules. In fact, despite the elimination of oxygen from the system inhibited HO. formation from H 2 O 2 photolysis, the production of hydroxyl radicals through the adsorbed water and/or hydroxyl groups with positive holes (r 1 and r 2 ) cannot be ruled out. During each run, the decrease of the concentration of cupric ions (Fig. 3) is accompanied by the precipitation of a purple solid which mixed with TiO 2 particles. According to what reported by some of the Authors in a previous paper [14], this solid could be Cu(). When Cu(II) is totally converted to Cu(), i.e. at 12 min for P25 Degussa sample and TiO 2 Aldrich pure anatase (Fig. 3, full diamonds and circles), no further consumption of benzyl alcohol (Fig. 1, full diamonds and circles) neither production of both benzaldehyde and benzoic acid is observed (Figs. 2a and b, full diamonds and circles). The possibility to reuse a certain copper amount for more runs has been directly investigated (Fig. 4). After 9 min of oxidation run, the lamp has been switched off and an oxygen stream has been fed to the reactor for 11 min. After about 6 min the Cu(II) concentration reaches approximately the same value as that at the beginning of the run whereas during this step no further consumption of the substrate is recorded. At this time, the purple colour disappeared and the solution returned to be white. The system thus obtained (at 2 min), has been used for a new oxidation experiment after a nitrogen gas bubbling to purge oxygen and switching on the lamp Effect of initial Cu(II) concentration Figs. 5a and b show the results obtained in oxidation runs at ph = 2. with the same TiO 2 sample (Aldrich, pure anatase, 2 mg/l) but at different initial concentrations of Cu(II), added as CuSO Fig. 4. Normalized concentration profiles for Cu(II) (circles) and benzyl alcohol (squares) with light on and nitrogen purge or light off and oxygen purge. [Benzyl alcohol] o = 1.5 mm. [Cu(II)] o = 1.16 mm. TiO 2 (Aldrich, pure anatase) = 2 mg/l. ph = 2.. T = 25 C. A higher initial concentration of Cu(II) results into a decrease of the system reactivity as can be easily verified by comparing the half-life time for the substrate which changed from 5 to 9 min for [Cu(II)] o equal to 1.12 mm and 2. mm respectively. Moreover, for both runs, the selectivities to benzaldehyde, evaluated at 15 min (72%) and 9 min (57%), do not change significantly from the values calculated by using the data shown in Figs. 1 and 2a. A possible explanation for these findings could be found supposing a change of the light absorption properties of the reacting solutions, with an increasing inner filter effect at increasing Cu(II) initial concentrations. UV absorption measurements have been thus performed on the reacting solutions in order to evaluate the capability of cupric solutions to absorb the lamp radiation at the adopted wavelengths (5, 313 and 366 nm). However, the values, estimated for the molar extinction coefficients of cupric aquocomplexes at ph = 2., allowed to rule out the possibility of existence of any inner filter effect due to these species (data not reported). The search of an explanation of the observed reduced reactivity of the oxidation of the alcohol at increasing Cu(II) concentrations revealed some difficulties which forced the attention on the fact that for the runs considered a parallel increase of sulphate ions had to be taken into account. That is, since the salt used to prepare the solutions is cupric sulphate, it is evident that any increase of Cu(II) resulted into one of sulphates and that it has been necessary to understand the effect on the system reactivity exerted by these species Effect of initial sulphate concentration Some photocatalytic tests have been thus carried out varying the initial sulphate concentration with different additions of Na 2 SO 4 salt (Fig. 6). The results obtained in these experiments indicate that sulphate ions exert a marked inhibiting effect on the photoactivity of TiO 2, by decreasing the oxidation rate of benzyl alcohol and its conversion at increasing the initial sulphate concentration (full symbols). This behaviour can be ascribed, as reported by others [24,29,], both for the reaction between sulphate species and the positive

5 R. Marotta et al. / Chemical Engineering Journal 172 (211) (a) (b).8.6 [Benzaldehyde]/[Benzyl alcohol]o Fig. 5. Effect of initial Cu(II) concentration: Benzyl alcohol consumption (a) and benzaldehyde production (b). TiO 2 (Aldrich, pure anatase) = 2 mg/l, ph = 2., T = 25 C. [Benzyl alcohol] o = 1.5 mm. [Cu(II)] o: 1.12 mm, 1.42 mm, 1.84 mm, 2. mm. holes (h + ) with the formation of sulphate ion radicals (SO 4 ) on the illuminated titanium dioxide surface: SO h + SO 4 (r3) and for a radical scavenging effect of sulphate ions: SO HO SO 4 + HO (r4) SO 4 species are reported to be less reactive than the hydroxyl radicals towards organic molecules [24]. In any case, as reported by Abdullah and coworkers [31], a catalyst deactivation, by the adsorbed sulphate ions which can block the TiO 2 active sites, cannot be ruled out. The selectivity to benzaldehyde which increased at increasing sulphate concentration (empty symbols) confirms the capability of sulphate ions to scavenge very reactive and unselective hydroxyl radicals Therefore, the results reported in Fig. 5 could not be attributed only to an effect of Cu(II) concentration but also to a combination of the latter and the concentration of sulphate ions. In particular, the negative effect of sulphate on the system reactivity prevailed over the effect exerted by Cu(II) Effect of TiO 2 load The effect of initial TiO 2 load was successively investigated by carrying out some oxidation experiments in which different amounts of the photocatalyst per litre were added to the reacting solutions (Figs. 7a and b). As expected, an increase of the catalyst load from 55 mg/l to 2 mg/l results into a marked increase of the system reactivity with the half-life time for the substrate ranging respectively from 24 min to 6 min. An increase of TiO 2 load, for values higher than 2 mg/l, does not lead to an increase of system reactivity probably due to scattering and screening of radiation by the excess particles, which mask part of the photosensitive surface [26,32] [Benzaldehyde]/[Bz Alcohol]o 3.5. Effect of ph Fig. 8a and b report the results obtained varying the ph of the solution, in the range During the tests, increasing the ph from 2. to 4., a decrease of benzyl alcohol consumption and benzaldehyde formation rates occurred. These results have two main causes. First of all, TiO 2 surface may be characterized by an amphoteric character, that is either positive or negative charge could be present on it as function of ph: TiOH + 2 TiOH + H + (r5) TiOH TiO + H + (r6) Fig. 6. Effect of initial sulphate concentration: [Benzyl alcohol] o = 1.5 mm. [Cu(II)] o = 1.15 mm. TiO 2 (Aldrich, pure anatase) = 2 mg/l. ph = 2.. T = 25 C. Full symbols: benzyl alcohol consumption, empty symbols: benzaldehyde production. [SO 2 4 ]o: (, ) 1.15 mm; (, ) 1.9 mm; (, ) 2.4 mm. The ph s value at which TiO 2 has a net zero surface charge is defined ph zpc. The surface has a net positive charge for ph < ph zpc, whereas for ph > ph zpc, the surface has a net negative charge. It could be supposed, according to the ph zpc equal to 4.2 for the adopted TiO 2 Aldrich sample [22], that an increase of the ph from 2. to 4. reduces the concentration of positive charges on the catalyst surface partially inhibiting the adsorption (and the reactivity) of benzyl alcohol (pk a = 15.2 [25]).

6 248 R. Marotta et al. / Chemical Engineering Journal 172 (211) (a) (b).8.6 [Benzaldehyde]/[Bz Alcohol]o Fig. 7. Effect of initial TiO 2 load. Benzyl alcohol consumption (a) and benzaldehyde production (b). [Benzyl alcohol] o = 1.5 mm. [Cu(II)] o = 1.5 mm. ph = 2.. T = 25 C. TiO 2 (Aldrich, pure anatase): 55 mg/l, 1 mg/l, 15 mg/l, 2 mg/l. On the other hand, since phosphoric acid has been used to adjust the ph of the reacting solutions, in the ph range 2. 4., the prevalent species present is H 2 PO 4. Since the capability of H 2 PO 4 ion to react with the positive holes formed on the catalyst after radiation is well known [23]: H 2 PO 4 + h + H 2 PO 4 (r7) it is clear that for any increase of ph resulting into a higher concentration of this ion, a marked inhibition of the direct oxidation of the substrate could be expected. In order to confirm these hypothesis, some further photo-oxidation tests have been performed, with and without an addition of NaH 2 PO 4 (2.44 mm) salt, on benzyl alcohol solutions (1.6 mm), at ph = 2. (regulated with H 3 PO 4 ), in the presence of TiO 2 (2 mg/l, Aldrich, pure anatase) and an initial [Cu(II)] equal to 1.15 mm (data not shown). After 12 min of reaction, when Cu(II) has been completely reduced to Cu(), benzyl alcohol total conversions of 55.% (without NaH 2 PO 4 ) and 45.8% (with NaH 2 PO 4 ) have been achieved respectively. Moreover, the percentage of the substrate converted into benzaldehyde has been 25.7% (without NaH 2 PO 4 ) and 32.7% (with NaH 2 PO 4 ), thus indicating a capability of dihydrogen phosphate ions to behave as a radical scavenging towards hydroxyl radicals [23]: H 2 PO 4 + HO H 2 PO 4 + HO (r8) Starting from previously discussed data, collected at different experimental conditions, the following simplified pictorial scheme of the mechanism for the selective photooxidation of benzyl alcohol to benzaldehyde and benzoic acid by TiO 2 photocatalysis could be depicted (Scheme 2). The irradiation of the photocatalytic surface leads to the formation of positive holes (h + ) in the valence band (VB) and electrons (e ) in the conduction band (CB). First of all, the positive holes 1. (a) (b).8.6 [Benzaldehyde]/[Bz Alcohol]o Fig. 8. Effect of ph. Benzyl alcohol consumption (a) and benzaldehyde production (b). [Benzyl alcohol] o = 1.5 mm. [Cu(II)] o = 1.5 mm. TiO 2 (Aldrich, pure anatase) = 2 mg/l. ph = 2., ph = 2.5 ph = 3., ph = 4..

7 R. Marotta et al. / Chemical Engineering Journal 172 (211) Scheme 2. Mechanism of selective oxidation of benzyl alcohol by TiO 2/Cu(II)/UV. react with benzyl alcohol (substrate) and benzaldehyde (intermediate) to produce benzaldehyde and benzoic acid respectively. Surface hydroxyl radicals are also formed by the reaction of water molecules with the holes. Hydroxyl radicals attack both benzyl alcohol and benzaldehyde to form undesired by-products (2- and 4- hydroxy benzyl alcohols, 2- and 4-hydroxy benzaldehyde). Finally, the holes can be trapped by sulphate and dihydrogenphosphate ions to generate less reactive SO 4 and H 2 PO 4 species. The electrons in the conduction band react with Cu(II) ions which are reduced to Cu(I) and Cu(). 4. Conclusion The possibility to oxidize benzyl alcohol to benzaldehyde in aqueous solution, under acidic conditions, using the photocatalytic system TiO 2 /Cu(II)/solar UV has been studied in the present work. Four samples of TiO 2 characterized by different crystallographic forms and specific surface areas have been used during the experiments. The best result found was a yield of 35% of benzaldehyde with respect of the initial benzyl alcohol. Benzaldehyde has been also partially converted to benzoic acid. The presence of undesired by-products, such as 2-hydroxy-benzylalcohol, 4-hydroxy-benzyl-alcohol, 2-hydroxy-benzaldehyde and 4-hydroxy-benzaldehyde, has been demonstrated and indicated an active production of surface HO radicals. At the end of the process copper (II) was totally reduced to copper (), which could easily be reoxidized to Cu(II), in the dark, in the presence of oxygen. The effect of the catalyst load, the nature of the inorganic anions (SO 2 4, H 2 PO 4 ), the initial Cu(II) concentration and ph of the solution has been also investigated. The sulphate and dihydrogen-phosphate anions resulted to exert a negative effect on the photooxidation rates of benzyl alcohol and to behave as scavengers towards surface HO radicals. A decrease of benzyl alcohol oxidation and benzaldehyde formation rates has been observed by increasing the ph from 2. to 4.. Acknowledgements The Authors are grateful to Ing. Luca Micoli for his assistance on BET measurements. References [1] K. Kabra, R. Chaudhary, R.L. Sawhney, Treatment of hazardous organic and inorganic compounds through aqueous-phase photocatalysis: a review, Ind. Eng. Chem. Res. 43 (24) [2] M. Kitano, M. Matsuoka, M. Ueshima, M. Anpo, Recent developments in titanium oxide-based photocatalysts, Appl. Catal. A: General 325 (27) [3] M.N. Chong, B. Jin, C.W.K. Chow, C. Saint, Recent developments in photocatalytic water treatment technology: a review, Water Res. 44 (21) [4] J.M. Herrmann, Heterogeneous photocatalysis: state of the art and present applications, Top. Catal. 34 (25) [5] Y. Shiraishi, T. Hirai, Selective organic transformations on titanium oxide-based photocatalysts, J. Photochem. Photobiol. C: Photochem. Rev. 9 (28) [6] S. Higashimoto, N. Kitao, N. Yoshida, T. Sakura, M. Azuma, H. Ohue, Y. Sakata, Selective photocatalytic oxidation of benzyl alcohol and its derivatives into corresponding aldehydes by molecular oxygen on titanium dioxide under visible light irradiation, J. Catal. 266 (2) (29) [7] S. Higashimoto, N. Suetsugu, M. Azuma, H. Ohue, Y. Sakata, Efficient and selective oxidation of benzylic alcohol by O 2 into corresponding aldehydes on a TiO 2 photocatalyst under visible light irradiation: Effect of phenyl-ring substitution on the photocatalytic activity, J. Catal. 274 (21) [8] R. Gao, A. Safrany, J. Rabani, Reactions of TiO 2 excess electron in nanocrystallite aqueous solutions studied in pulse and gamma-radiolytic systems, Radiat. Phys. Chem. 67 (23) [9] M.I. Litter, Heterogeneous photocatalysis transition metal ions in photocatalytic systems, Appl. Catal. B: Environ. 23 (1999) [1] G. Palmisano, S. Yurdakal, V. Augugliaro, V. Loddo, L. Palmisano, Photocatalytic selective oxidation of 4-methoxybenzyl alcohol to aldehyde in aqueous suspension of home-prepared titanium dioxide catalyst, Adv. Synth. Catal. 349 (27) [11] V. Augugliaro, T. Caronna, V. Loddo, G. Marcì, L. Palmisano, S. Yurdakal, Oxidation of aromatic alcohols in irradiated aqueous suspensions of commercial and home-prepared rutile TiO 2: a selectivity study, Chem. Eur. J. 14 (28) [12] S.W. Zou, C.W. How, J.P. Chen, Photocatalytic treatment of wastewater contaminated with organic waste and copper ion from semiconductor industry, Ind. Eng. Chem. Res. 46 (27) [13] K. Kabra, R. Chaudhary, R.L. Sawhney, Effect of ph on solar photocatalytic reduction and deposition of Cu(II), Ni(II), Pb(II) and Zn(II): speciation modelling and reaction kinetics, J. Hazard. Mater. 149 (27) [14] M. Canterino, I. Di Somma, R. Marotta, R. Andreozzi, Kinetic investigation of Cu(II) ions photoreduction in presence of titanium dioxide and formic acid, Water Res. 42 (28) [15] N.S. Foster, R.D. Noble, C.A. Koval, Reversible photoreductive deposition and oxidative dissolution of copper ions in titanium dioxide aqueous suspensions, Environ. Sci. Technol. 27 (2) (1993) [16] N.S. Foster, A.N. Lancaster, R.D. Noble, C.A. Koval, The effect of organic hole scavenger on the photodeposition of copper in titanium dioxide suspensions, Ind. Eng. Chem. Res. 34 (1995) [17] J.W.M. Jacobs, F.W.H. Kampers, J.M.G. Rikken, C.W.T. Bulle-Lieuwma, D.C. Koningsberger, Copper photodeposition on TiO 2 studied with HrEM and EXAFS, J. Electrochem. Soc. 136 (1989) [18] M. Bideau, B. Claudel, L. Faure, M. Rachimoellah, Photooxidation of formic acid by oxygen in the presence of titanium dioxide and dissolved copper ions: oxygen transfer and reaction kinetics, Chem. Eng. Commun. 93 (199) [19] S. Morishita, Photoelectrochemical deposition of copper on TiO 2 particles. Generation of copper patterns without photoresists, Chem. Lett. (1992) [2] Ullmann s Encyclopedia of Industrial Chemistry, sixth ed. Wiley-VCH, Weinheim, 25. [21] G. Fattima Al-Zahra Gassima, N. Ahmed Alkhateebb, H. Falah Hu, Photocatalytic oxidation of benzyl alcohol using pure and sensitized anatase, Desalination 29 (27) [22] J. Ryu, W. Choi, Substrate-specific photocatalytic activities of TiO 2 and multiactivity test for water treatment application, Environ. Sci. Technol. 42 (28) 294. [23] S. Ahmed, M.G. Rasul, N. Martens Wayde, R. Brown, M.A. Hashib, Heterogeneous photocatalytic degradation of phenols in wastewater: a review on current status and developments, Desalination 261 (21) [24] C. Hu, T. Yuchao, L. Lanyu, H. Zhengping, W. Yizhong, T. Hongxiao, Effects of inorganic anions on photoactivity of various photocatalysts under different conditions, J. Chem. Technol. Biotechnol. 79 (24) [25] F.M. Menger, M. Ladika, Origin of rate accelerations in an enzyme model: the p-nitrophenyl ester syndrome, J. Am. Chem. Soc. 19 (1987) [26] J.M. Hermann, Heterogeneous photocatalysis: state of the art and present applications, Top. Catal. 34 (1 4) (25) [27] R. Long, Y. Dai, B. Huang, Structural and electronic properties of iodine-doped anatase and rutile TiO 2, Comput. Mater. Sci. 45 (29) [28] H.A. Schwarz, R.W. Dodsonj, Equilibrium between hydroxyl radicals and thallium(ii) and the oxidation potential of OH(aq), J. Phys. Chem. 88 (1984) [29] O. Carp, C.L. Huisman, A. Reller, Photoinduced reactivity of titanium dioxide, Prog. Solid State Chem. 32 (24) [] M. Ziegmann, T. Doll, F.H. Frimmel, Matrix effects on the photocatalytical degradation of dichloroacetic acid and atrazine in water, Acta Hydrochim. Hydrobiol. 34 (26) [31] M. Abdullah, G.K.C. Low, R.W. Matthews, Effects of common inorganic anions on rates of photocatalytic oxidation of organic carbon over illuminated titanium dioxide, J. Phys. Chem. 94 (199) [32] D. Vione, C. Minero, V. Maurino, M.E. Carlotti, T. Picatonotto, E. Pelizzetti, Degradation of phenol and benzoic acid in the presence of a TiO 2-based heterogeneous photocatalyst, Appl. Catal. B: Environ. 58 (25)