Catalytic degradation of Acid Orange 7 by H2O2 as promoted by either bare or V-loaded titania under UV light, in dark conditions, and after incubating the catalysts in ascorbic acid

Abstract Pure and V-loaded mesoporous titania (with 2.5 wt-% V) were prepared by template-assisted synthesis and compared to commercial titania (Degussa P25), both as such and after vanadium loading. Mesoporous TiO2 occurred as pure anatase nanoparticles with higher surface area (SSA = 150 m2 g−1) than P25 (SSA = 56 m2 g−1). Degradation of the azo dye Acid Orange 7 by H2O2 was used as a test reaction: under UV light, no difference emerged between mesoporous TiO2 and P25, whereas in dark conditions, higher SSA of the mesoporous sample resulted in higher conversions. Under UV illumination, surface V5+ species inhibited photocatalytic activity, by forming inactive V4+ species. Similarly, in dark conditions, V5+ surface species reacted with H2O2, likely yielding ·O2H radicals and reducing to V4+. On the contrary, V-containing catalysts were very active after pretreatment with ascorbic acid, which reduced V5+ species to V3+species, the latter promoting very lively a Fenton-like reaction.


Introduction
Titania, anatase in particular, is both a very popular photocatalyst and catalytic support, 1-7 not only because of its (reasonably small) band gap, but also for the possibility of obtaining it in nanoporous and/or nanoparticle form, with increased specific surface area (SSA). [8][9][10][11][12] One current application of TiO 2 is the degradation of organic pollutants, including azo dyes, 4 a class of organic molecules widely applied in the photographic industry (and bearing an environmental impact once released in the surroundings), 13 which can be removed by photocatalytic degradation with TiO 2 , usually in the presence of an oxidizing agent, like H 2 O 2 . 2 P25 is one of the most common commercial forms of TiO 2 , occurring as a mixture of mainly anatase and rutile with SSA = 50 ± 15 m 2 g −1 . 14 In the present paper, high SSA mesoporous TiO 2 (MT) was obtained by template-assisted synthesis, and calcined at 450 °C to avoid phase transition to rutile. Anatase has, indeed, a larger adsorptive affinity than rutile for organic molecules 15 and is generally regarded as the most photocatalytically active phase of TiO 2 .
Since vanadium has been shown to improve the catalytic activity of TiO 2 , extending its absorption the Vis-range, 16 a sample of MT with a 2.5 wt-% V loading was prepared; for comparison, two samples were obtained by impregnating P25 with (i) same V content and (ii) 1/3 V content to get, in principle, the same vanadium dispersion as for V2.5-MT.
The degradation of a model azo dye (Acid Orange 7, AO7) 13 was studied either under UV illumination or not (hereafter referred to as "dark conditions"). The effects of H 2 O 2 concentration in the reaction and the presence of ascorbic acid (H 2 Asc) were also studied. H 2 Asc, especially in biochemical processes, plays different roles as an antioxidant, a "mild" reductant of both metal ions and organic moieties, and a radical scavenger. 17 It is able, inter alia, to reduce transition metal ions, including vanadium, in biological systems. 18,19 Nonetheless, reports are available on the effects of an oxidant-reductant system (H 2 O 2 and H 2 Asc, for instance) in the catalytic activity of V-containing catalysts. 20,21 Original Research Paper Catalytic degradation of Acid Orange 7 by H 2 O 2 as promoted by either bare or V-loaded titania under UV light, in dark conditions, and after incubating the catalysts in ascorbic acid

Experimental Catalysts preparation
All reagents were ACS grade chemicals from Sigma-Aldrich.
The synthesis of mesoporous titania (MT) and of V-containing MT with a vanadium nominal content of 2.5 wt-% (V2.5-MT) is detailed in Ref. 23. Vanadium resulted present mainly at the surface of V2.5-MT. 22 The samples were calcined at 450 °C in air to remove the template and to avoid phase transition to rutile.
The commercial titania (CT) was Degussa P25 (TiO 2 content ≥99.5%); two samples with nominal V content of either 2.5 or 0.80 wt-% (V2.5-CT and V0.80-CT) were obtained by impregnation with NH 4 VO 3 solution, followed by drying at 60 °C and calcination in air at 450 °C for 4 h.

Catalysts characterization
Powder X-ray diffraction patterns were collected on a X'Pert Philips PW3040 diffractometer using Cu Kα radiation (2θ range = 20°-85°; step = 0.05° 2θ; time per step = 0.2 s) and indexed according to the Powder Data File database (PDF 2000, International Centre of Diffraction Data, Pennsylvania). Crystallites average size (D) was determined by using the Debye-Scherrer formula, D = 0.9 λ/b cosθ, where λ is the wavelength of the Cu Kα radiation, b is the full width at half-maximum (in radians), 0.9 is the shape factor for spherical particles, and θ is the angle of diffraction peaks. The anatase content was evaluated by the full-profile Rietveld method applied to diffraction patterns using the GSAS-EXPGUI free software. XRD background was modeled by a 10-term cosine polynomial function, and pseudo-Voigt functions were adopted for peaks curve fitting.
The SSA and total porous volume (V p ) were measured by N 2 sorption isotherms at −196 °C (Quantachrome Autosorb 1C) on powders outgassed at 150 °C for 4 h to remove water and other atmospheric contaminants; the SSA was determined according to the Brunauer-Emmett-Teller method.
The metal content was determined by (semi-quantitative) chemical analysis carried out by means of an energy-dispersive X-ray probe (low-vacuum Scanning Electron Microscope Quanta inspect 200) on 10-50-nm diameter spots. For each sample, about 10 measurements were carried out in different spots of the sample, from which an average metal content was calculated, as reported in Table 1.
Diffuse reflectance (DR) UV-vis spectra of powder samples dehydrated at 150 °C were measured on a Cary 5000 UV-vis-NIR spectrophotometer (Varian instruments) equipped with a DR sphere. X-ray photoelectron spectroscopy (XPS) measurements were obtained on an XPS PHI 5000 Versa probe apparatus. The C 1s peak at 284.6 eV was used as a reference for charge correction. To study spent catalysts, powders recovered by centrifugation were dried in air and pressed in order to obtain pellets that were further outgassed under vacuum at room temperature before XPS analysis.

Catalytic tests
A total volume of 50 mL of 0.67 mM aqueous solution (natural pH of 6.80) of AO7 (Fluka) and an amount of catalyst corresponding to 1.0 g L −1 concentration were used systematically for catalytic tests, during which the suspension was stirred by means of a magnetic stirrer, operated at 600 rpm. Aliquots of the suspension were collected at regular intervals of time, the supernatant fraction was separated by centrifugation (ALC centrifuge PK110, at 4000 rpm for 2 min) and the UV-vis spectrum was measured in the 190-800 nm range on a Cary 5000 UV-vis-NIR spectrophotometer (Varian instruments), using a quartz cell with 1-mm path length. The concentration of AO7 was evaluated by the intensity of its band at 484 nm (vide infra), after a proper calibration procedure. (1)

Survey of the main physicochemical features of the catalysts
The XRD analysis (not reported) showed that anatase is the only phase present in MT samples (99.8%, as obtained by Rietveld refinement), whereas P25 is a mixture of anatase (88.8%) and rutile (11.2%). The size of MT crystallites was around 15 nm, as calculated according to the Debye-Scherrer formula (Table 1), in agreement with previous transmission electron microscopy observation. 22 Type IV N 2 isotherms were measured at −196 °C on MT samples, with high SSA values due to both intra-and interparticle mesopores; a limited decrease of SSA was observed with V2.5-MT (Table 1). Fig. 1 reports DR UV-vis spectra of samples outgassed at 150 °C to remove water and other atmospheric contaminants: comparison between MT and P25 shows that the former absorbs in a broader range of wavelengths (Fig. 1a). The spectrum of MT is shifted toward higher wavelengths with respect to P25 (arrow): accordingly, the corresponding Tauc's plots, in the version for indirect band gap (E g ) semiconductors (inset to Fig. 1a), yield E g ≈ 3.30 eV for P25 and E g ≈ 3.15 eV for MT. The redshift observed in MT absorption edge is likely due to slightly different optical and electrical properties of the mesoporous material with respect to bulk TiO 2 , as already observed with porous films of titania. 23 Fig. 1b compares the UV-vis spectra of V-containing catalysts, characterized by a pale yellow color: with respect to parent MT, V2.5-MT shows a small redshift of the absorption band and an increased absorption above 375 nm. Both isolated VO x and oligomeric V x O y species (y = 2x + 1 in dehydrated powders), respectively, absorbing at 270 and 325 nm, 24,25 are likely present at the surface, though their occurrence is masked by TiO 2 absorption. V2.5-CT has similar UV-vis spectrum, whereas absorption related to VO x species is less intense with V0.80-CT, in agreement with the lower metal content.

Preliminary considerations about catalytic tests
Scheme 1 reports the possible structures of AO7 in water: the hydrazone form (B), stable in the solid phase, undergoes, in water, azo-hydrazone tautomerism via intramolecular proton transfer, so that both hydrazone (B) and azo-form (A) are simultaneously present in solution.
The UV-Visible spectrum of 0.67 mM AO7 in water is reported as the bold curve in Fig. 2: the two peaks at 310 and 230 nm and the shoulder at 256 nm are due to aromatic ring absorptions. The peak at 484 nm is due to the n-π* transition involving the lone pair on the N atoms and the conjugated system extending over the two aromatic moieties and encompassing the N-N group of the hydrazone form. 26 The shoulder at 403 nm has a similar nature, involving the N-N group of the azo-form. 26 The (non-catalytic) reaction between AO7 and H 2 O 2 (either 0.030 M or 0.80 M) was preliminarily studied. The UV-vis spectra taken with 0.80 M H 2 O 2 are reported in Fig. 2: the region below 300 nm is dominated by H 2 O 2 absorption; small changes are observed in the AO7 bands, for which a limited decrease in intensity is observed after 24 h, as H 2 O 2 is, per se, able to attack aromatic rings. 27 This shows that without catalyst, a limited conversion of AO7 is obtained, even in the presence of excess H 2 O 2 . 186 photocatalytic activity of V-doped TiO 2 under UV light, though extending absorption in the Vis range. 16 Results obtained under UV light with 0.030 M H 2 O 2 are reported in Fig. 4: as expected, addition of H 2 O 2 to the reaction mixture had a positive effect on the photocatalytic activity of all the samples. With pure TiO 2 (Fig. 4a), H 2 O 2 most probably acts as an acceptor of photogenerated electrons (e − ), according to reaction (4):

Catalytic tests under illumination
forming hydroxyl radicals for the degradation of AO7. Again, MT and P25 show the same activity as before, and V has almost no effect on photocatalytic activity (Fig. 4b), which seems only due to TiO 2 .

Catalytic behavior in dark conditions
Successive tests were run without UV irradiation to exploit the mere surface properties of the catalysts. Fig. 5a reports, as an example, spectra obtained after contacting AO7 solution with MT sample and 0.030 M H 2 O 2 in dark conditions; Fig. 5b reports results obtained in dark conditions with the same catalyst and 0.80 M H 2 O 2 . In both cases, bands of AO7 decreased in intensity: no new band, related to any decomposition product, however, appeared. As shown in Fig. 2, the spectral region below 300 nm is dominated by H 2 O 2 absorption at higher concentration of the reactant: this feature was observed with all catalysts, so that the presence, if any, of reaction products absorbing in this region was not detectable. On the other hand, the intensity of the absorption below 340 nm does not decrease, so showing that the 0.80 M solution provides excess H 2 O 2 , in agreement with previous stoichiometric considerations.   29 and so reaction (5) may slow down AO7 depletion: With respect to P25, conversion reached with MT is higher, in agreement with the higher SSA of the latter catalyst; concerning initial rate, the positive effect of MT SSA may be appreciated at 0.80 M H 2 O 2 concentration ( Table 2).
Reactivity of Ti with H 2 O 2 is well documented in the literature as it concerns Ti-Silicalite, and involves formation of (4) an acceptor of e − (equation (2)) and the powerful oxidant h + reacts with OH − ions, generating radical HO· species (equation (3)) taking part in the AO7 degradation: The MT and P25 had the same photocatalytic activity (Fig.  3a), whereas V2.5-CT and V2.5-MT (Fig. 3b) were practically inactive. The same behavior of the bare TiO 2 samples, notwithstanding the different SSA, may be surprising. It has to be recalled, however, that P25, an optimized commercial sample, contains a substantial fraction of rutile, which is recognized to be able to prevent the fast electron-hole recombination occurring in anatase, due to a positive synergic effect between anatase and rutile nanoparticles in P25. 14,28 The inactivity of V-containing TiO 2 has another explanation: surface V 5+ centers may trap electrons, giving rise to reduced (and probably inactive) V 4+ species. Inhibition of TiO 2 reactivity by surface V species is probably related to the fact that the latter act as recombination centers for electron-hole pairs, eventually suppressing the intrinsic photocatalytic activity of the support. This is in agreement with previous literature results, showing that vanadium has a detrimental effect on the The "quenching effect" of H 2 O 2 observed at higher concentration (Table 2) is in agreement with the hypothesis of a radical mechanism, in which reactive OH· species are, in this case, due to decomposition of Ti-OOH. Table 2 shows that with 0.030 M H 2 O 2 , initial rates are higher in the presence of V2.5-MT and V2.5-CT catalysts, whereas with 0.80 M H 2 O 2 V-containing samples are very active at the peroxo bridges with expansion of the coordination sphere of Ti. 30 Concerning TiO 2 , in the presence of H 2 O 2 , surface Ti-OH groups are partially converted into Ti-OOH species, more reactive than H 2 O 2 in partial oxidation reactions. 31 Fig. 7 reports XPS spectra of MT before and after reaction with 0.80 M H 2 O 2 , for Ti 2p (section a) and O 1s range (section b). Before reaction, with MT, two peaks are seen at 464.02 and 458.28 eV, respectively, assigned to the 2p 1/2 and 2p 3/2 lines of Ti 4+ , 32 the spectrum of P25 being very similar. After reaction, MT XPS spectrum shifts to higher BE values: the same behavior, observed for anatase treated with H 2 O 2 , 32 was ascribed to the formation of surface Ti-OOH groups, as the peroxyl group has an electron-withdrawing effect and causes a shift to higher BE values. 32 The O 1s region of MT and P25 XPS spectra is similar before reaction: that of MT shifts to higher BE values after reaction. This could be ascribed to a more extensive formation of Ti-OOH species in MT, due to the higher SSA. Curve fits of the O 1s range are reported for MT in Fig. 7c and d: before reaction, a satisfactory curve fit was obtained with two peaks at 529.53 and 530.90 eV, due to O 2− related to Ti 4+ and to OH − species, respectively. Conversely, after reaction, a satisfactory curve fit was obtained with three peaks, the additional peak at 532.33 eV being ascribed to Ti-OOH species. 32 Although based on delicate curve fitting procedures, such result is in agreement with previous literature reports on H 2 O 2 -treated 188 and a sudden initial increase and so the activity of doped surfaces should thus be not only to metal sites, but also to such patches.
In agreement with the literature, 33,34 we propose that reduction of V 5+ to V 4+ species occurs according to the reactions (6) and (7): beginning of the reaction, then conversion reaches a plateau (Fig. 6), indicating that they are subjected to deactivation. The surface density of heteroatoms (Table 1)  V 4+ is formed together with a hydroperoxyl radical HO 2 · than can react with excess H 2 O 2 forming more reactive HO· (equation (8)). Such process is not catalytic, but stoichiometric, and indeed vanadium effect is only observed at the beginning of the reaction, the final conversion being almost the same as with bare TiO 2 . If the MT conversion curve is subtracted from that of V2.5-MT, indeed, only 10% AO7 results to be converted by surface vanadium species, indicating that a limited fraction of surface V sites is reactive.
Formation of surface V 4+ species is confirmed by XPS measurement run V2.5-MT: before reaction, two peaks are seen at   (Fig. 11b); the same effect was observed with both V2.5-CT and V0.80-CT.
V 5+ surface species are likely reduced by H 2 A to V 3+ species, which may undergo Fenton-like 41,42 reaction switching to V 4+ : In this case, the presence of an excess of H 2 Asc ensured reduction of all surface species and allowed several redox cycles, in which very reactive V 3+ species were regenerated, avoiding catalyst deactivation due to formation of V 4+ .

Conclusions
In dark conditions, MT is more active than P25 in the degradation of AO7, likely due to its larger SSA, which favors the formation of reactive Ti-OOH species in the presence of H 2 O 2 ; under UV illumination, the two types of TiO 2 showed comparable activity, due to a balance between higher SSA of MT and mixed crystalline composition of P25.
Absorption of UV light gives rise to the formation of inactive V 4+ species, and therefore, surface doping with V may be detrimental to photocatalytic activity. In dark conditions, V 5+ species are reduced by H 2 O 2 to the same inactive V 4+ ions, but after contact with H 2 Asc, reduction to V 3+ species occurs, the latter species being, instead, very active in a Fenton-like mechanism.

Acknowledgments
AO7 + HO⋅ → degradation products 515.9 and 523.4 eV, due to pentavalent V 2p 3/2 and V 2p 1/2 lines, respectively (Fig. 8). 35 The shift in peaks position after reaction shows that V 5+ species underwent reduction to V 4+ , in agreement with the literature reports on V-doped TiO 2 , [36][37][38] supporting the mechanism proposed. Moreover, in both cases V 2p 3/2 lines were satisfactorily curve-fitted with one peak, so showing that practically all the surface V 5+ species present before reaction were reduced to V 4+ .

Effect of ascorbic acid on catalytic activity in dark conditions
H 2 Asc, a well-known reducing agent, was used to reduce vanadium to V 3+ species, which are active in the Fenton-like reaction (vide infra).
Blank experiments without any catalyst were firstly run by contacting 0.67 mM AO7 solution with either 0.40 M H 2 Asc or 0.80 M H 2 O 2 , and with both chemicals (0.40 M H 2 Asc and 0.80 M H 2 O 2 ). Fig. 9 compares the related UV-vis spectra: curve 2 is obtained with 0.40 M H 2 Asc after 120-min contact, showing that some reaction occurs between AO7 and H 2 Asc, whereas the strong absorption below 320 nm is due to H 2 Asc itself. 39 Fig. 11 compares catalytic activities in dark conditions with and without H 2 Asc. H 2 Asc depresses the activity of MT; the conversion remaining constant to a value well below that obtained without H 2 Asc (Fig. 11a). In agreement with the Scheme 1 Azo-and hydrazone form of AO7