Photocatalytic Activity vs Structural Features of Titanium Dioxide

Oct 8, 2014 - Photocatalytic Activity vs Structural Features of Titanium Dioxide Materials Singly Doped or Codoped with ... *Phone: +39 02 50314237...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/JPCC

Photocatalytic Activity vs Structural Features of Titanium Dioxide Materials Singly Doped or Codoped with Fluorine and Boron Maria Vittoria Dozzi,† Luca Artiglia,‡ Gaetano Granozzi,‡ Bunsho Ohtani,§ and Elena Selli*,† †

Dipartimento di Chimica, Università degli Studi di Milano, via Golgi 19, I-20133 Milano, Italy Dipartimento di Scienze Chimiche, Università degli Studi di Padova, via Marzolo 1, I-35131 Padova, Italy § Catalysis Research Center, Hokkaido University, Sapporo 001-0021, Japan ‡

ABSTRACT: TiO2 photocatalysts, either doped with different amounts of fluorine or boron, or codoped with boron and fluorine, were prepared by sol−gel synthesis, followed by calcination at different temperatures (500−700 °C), and characterized by XRPD, BET, XPS, and UV−vis absorption analyses. The oxidative decompositions of both formic and acetic acid were employed as photoactivity test reactions, also in comparison with previous results obtained with NF-codoped TiO2. A detailed XPS analysis revealed the presence of B2O3 on the surface of B-doped and BF-codoped TiO2, which could be removed by washing under acidic conditions and had no beneficial role in phototocatalysis. A correlation was found between the photoactivity of full anatase TiO2 materials calcined at 500 °C, singly doped or codoped with nitrogen, fluorine, and boron, and their specific surface area, pointing to a major role of the latter, not of the different dopants, on photoactivity. Only samples containing fluorine, as dopant or codopant, exhibit a photoactivity increase with increasing the calcination temperature to 700 °C. This fact, together with an action spectra analysis on photocatalysts calcined at different temperatures, evidenced that only fluorine is responsible for the photoactivity increase in the UVA region observed with full-anatase, highly crystalline doped TiO2 calcined at high temperature. production from water/methanol vapors.5−8 Aiming at ascertaining the activation of doped materials under visible light, the photocatalytic oxidation of acetic acid was also investigated as a function of the irradiation wavelength through an action spectra analysis.6 By comparing the shapes of these latter with those of the absorption spectra of the investigated semiconducting materials, a clear distinction could be outlined between absorption of light leading to no photocatalytic activity and irradiation wavelength-dependent effective photoactivity.6 The role of fluorine and of nitrogen as codopants in TiO2 has also been clarified by comparing the absorption features with the photoactivity of TiO2 materials prepared by the same synthetic procedure, employing either HF or NH4F as dopant source.8 The effect of the copresence of p-block elements boron and fluorine in titania has now been explored, together with those of single boron doping. The use of boron as TiO2 dopant to impart photoactivity under visible light needs indeed to be clarified, previous studies on such doped materials being unable to demonstrate their photoactivity under visible light. In fact, organic dyes were mainly employed as degradation substrates9−11 and the activity results obtained under such conditions cannot be interpreted in a straightforward way, due to the possible involvement of a dye-sensitized mechanism.12,13

1. INTRODUCTION Several different routes have been explored in recent years to minimize the major drawback of titania (TiO 2) as a photocatalytic material, i.e., its high-energy band gap (3.2 eV). Indeed, TiO2 mainly absorbs in the UV region, and consequently, only a small portion (less than 5%) of the solar spectrum can be exploited for photocatalytic processes employing this highly stable and robust semiconductor photocatalyst. This represents a great limitation in its use in photocatalysis, particularly for solar into chemical energy conversion, i.e., in the production of so-called solar fuels. Anion doping with p-block elements has been widely investigated in the past decade to extend the action spectrum of TiO2 toward visible light.1−3 Either newly created midgap energy states can be introduced by this way or the semiconductor band gap itself can be narrowed. However, the nature and possible beneficial effects of doping titanium dioxide are still under lively debate. In fact, the insertion of dopant impurities in the oxide structure may increase the rate of the undesired recombination of photogenerated charge carriers, an effect becoming relatively lower the higher is the crystallinity of the oxide structure, and may induce visible light absorption by doped TiO2, though without guaranteeing its photoactivity in the visible region.4 The photocatalytic features of a large series of singly fluorinedoped and N,F-codoped TiO2 photocatalysts, mainly prepared by sol−gel synthesis, has been thoroughly investigated in previous studies, employing both liquid and gas-phase test reactions, including formic and acetic acid oxidation in the aqueous phase, acetaldehyde decomposition, and hydrogen © 2014 American Chemical Society

Received: August 21, 2014 Revised: October 4, 2014 Published: October 8, 2014 25579

dx.doi.org/10.1021/jp5084696 | J. Phys. Chem. C 2014, 118, 25579−25589

The Journal of Physical Chemistry C

Article

sulfate as a reference, and then converted into absorption (A) spectra (A = 1 − R). X-ray photoemission spectroscopy (XPS) data were collected in a multi-technique ultra-high-vacuum (UHV) chamber (base pressure: 1.0 × 10−9 mbar) equipped with a VG MKII ESCALAB electron analyzer (5 channeltrons). XPS data were taken at room temperature in normal emission using a nonmonochromatized Mg anode X-ray source (hν = 1253.6 eV). The calibration of the binding energy (BE) scale was determined using Au 4f as reference. Powder samples were suspended in bi-distilled water and drop-casted on high-purity copper foils. After drying in air, the obtained films were introduced into the UHV chamber and outgassed overnight. The charging observed during measurements was corrected by aligning the Ti 2p3/2 core-level peak signal to 459.0 eV. Sample sputtering was performed by means of an Omicron ISE 5 ion gun (1.0 keV Ar+ plasma for 5 min). 2.3. Formic Acid Photocatalytic Decomposition Tests. All formic acid photocatalytic degradation runs were performed under atmospheric conditions using the already described photoreactor and setup,15 consisting of an Osram Powerstar 150 W lamp, emitting ultraviolet and visible light at λ > 340 nm, mounted on an optical bench. The irradiated aqueous suspensions always contained 0.1 g L−1 of photocatalyst and a formic acid initial concentration equal to 1.0 × 10−3 mol L−1. The adsorption equilibrium of the substrate on the photocatalyst surface was attained under magnetic stirring, before starting irradiation. Stirring was continued during the photocatalytic runs. Portions (2 mL) of the suspension were withdrawn from the photoreactor at different time intervals during the runs and centrifuged employing an EBA-20 Hettich centrifuge. The supernatant was analyzed for residual formic acid content by ion chromatography with conductivity detection, employing a Metrohm 761 Compact IC instrument, after calibration for formate ion concentration in the 0−50 ppm range.15 2.4. Acetic Acid Oxidative Decomposition under Polychromatic Irradiation. The photocatalytic tests were performed in ca. 35 mL volume sealed glass tubes, as already described.6 A 50 mg portion of each photocatalyst powder was suspended in 5.0 mL of an aqueous solution containing 5.0 vol % of acetic acid. This amount of photocatalyst powder in the suspension was large enough to ensure complete absorption of the incident light. The suspensions were irradiated using an Eiko-sha 400-W high-pressure mercury lamp (emission wavelength > 290 nm). The reaction temperature was kept at 25 °C using a thermostated water bath. At regular time intervals during the runs, 0.2 mL samples of the gas in the tube were withdrawn with a gastight syringe and analyzed using a Shimadzu GC-8A gas chromatograph with a TC detector, equipped with MS-5A and Porapak-Q columns. The molar amount of evolved CO2 was calibrated considering the increase in pressure in the reaction tube consequent to the increased amount of gas-phase molecules. The reproducibility of the 80 min long kinetic runs was checked by repeating them at least twice. 2.5. Acetic Acid Decomposition − Action Spectra Analysis. Action spectra analysis was performed employing the photocatalytic decomposition of acetic acid in aerated liquid suspensions as a test reaction,6 by measuring the reaction rate as a function of irradiation wavelength in ca. 10.5 mL quartz cells. Each photocatalyst powder (20 mg) was suspended in an aqueous solution (2.0 mL) containing 5.0 vol % of acetic acid.

The photoactivity of TiO2 samples singly doped, either with fluorine or with boron, or also codoped with these two elements (BF-codoped TiO2) has been here investigated in the photooxidation of transparent formic acid and acetic acid in aqueous suspensions. Moreover, an action spectra analysis has been performed with three selected series of doped TiO2 full anatase samples. The aim of this study was to enlighten how the wavelength-dependent photocatalytic activity of the differently doped TiO2 materials can be correlated to their structural features depending on the dopant type and concentration and on the calcination temperature.

2. EXPERIMENTAL SECTION 2.1. Preparation of Doped TiO2. All photocatalysts were prepared by the already described sol−gel method,5,6 starting from an anhydrous ethanol solution containing a fixed concentration of titanium isopropoxide and different amounts of dopant source (HF for F-doped TiO2 and H3BO3 for Bdoped TiO2), in order to obtain dopant/Ti percent nominal molar ratios equal to 3, 5, 12, and 25. BF-codoped materials were prepared in the presence of the two dopant sources HF and H3BO3, so as to obtain dopant/Ti percent nominal molar ratios equal to 3, 5 and 12, with the HF/H3BO3 molar ratio always fixed at 1. The effect of the calcination temperature (at 500, 600, and 700 °C) on photoactivity was also investigated systematically. The so-obtained materials were labeled as X_Y_Z, with X referring to the dopant element symbol, Y to the nominal dopant/Ti percent molar ratio, and Z to the calcination temperature (°C). Two nitrogen-doped TiO2 samples were also prepared in the presence of 5 mol % NH3, as nitrogen dopant source, under otherwise identical conditions, and calcined at 500 and 700 °C (N_5_500 and N_5_700). Reference undoped materials, prepared by exactly the same synthetic route in the absence of a dopant source, are referred to as the T_0_Z photocatalysts series. After calcination, selected samples were washed under acidic conditions. A 30 μL portion of concentrated HCl (37%) was added to a suspension containing 500 mg of photocatalyst in 100 mL of water. The acid suspension, after being sonicated for 30 min, was maintained under constant stirring for 2 h. Then, the powder was separated by centrifugation and washed at least four times with 100 mL portions of water. After having checked, by ion chromatographic analysis, that the residual amount of chloride ions in the supernatant was below 1 ppm, the so-obtained washed material was dried in air at 70 °C for one night. All reagents were purchased from Aldrich and employed as received. Water purified by a Milli-Q water system (Millipore) was used throughout. 2.2. Characterization of Doped TiO2 Materials. The Brunauer−Emmett−Teller (BET) specific surface area (SSA) was measured by N2 adsorption/desorption at liquid nitrogen temperature in a Micromeritics Tristar II 3020 V1.03 apparatus, after outgassing at 300 °C for 1 h under a N2 stream. X-ray powder diffraction (XRPD) patterns were recorded on a Philips PW3020 powder diffractometer, using Cu Kα radiation (λ = 1.54056 Å). Quantitative phase analysis was made by the Rietveld refinement method,14 using the “Quanto” software. Diffuse reflectance (DR) spectra of the photocatalyst powders were recorded on a Jasco V-670 spectrophotometer equipped with a PIN-757 integrating sphere, using barium 25580

dx.doi.org/10.1021/jp5084696 | J. Phys. Chem. C 2014, 118, 25579−25589

The Journal of Physical Chemistry C

Article

Table 1. Phase Composition and Anatase Crystallite Dimensions (dA) Obtained from XRD Analysis Using the Scherrer Equation, by Assuming the Absence of Amorphous Phase; Specific Surface Area (SSA) Obtained from BET Analysis sample

dopant/Ti molar ratio (%)

anatase (%)

brookite (%)

T_0_500 T_0_600 T_0_700 F_3_500 F_3_600 F_3_700 F_5_500 F_5_600 F_5_700 F_12_500 F_12_600 F_12_700 F_25_500 F_25_600 F_25_700 B_5_500 B_5_600 B_5_700 B_12_500 B_12_600 B_12_700 B_25_500 B_25_600 B_25_700 BF_3_500 BF_3_600 BF_3_700 BF_5_500 BF_5_600 BF_5_700 BF_12_500 BF_12_600 BF_12_700 N_5_500 N_5_700

0 0 0 3 3 3 5 5 5 12 12 12 25 25 25 5 5 5 12 12 12 25 25 25 3 3 3 5 5 5 12 12 12 5 5

99 90 38 100 100 100 100 100 100 100 100 100 100 100 100 97 100 64 96 100 97 87 100 100 100 100 98 100 100 95 100 100 98 100 19

1

rutile (%) 10 62

3 36 4 3 13

2

5

2 81

dA (nm)

SSA (m2 g−1)

15 32 65 23 41 56 22 40 55 20 50 57 20 49 85 10 16 43 8 13 40 6 19 61 12 19 36 12 18 29 11 17 14 12 45

15 9 7 44 25 20 45 28 22 34 16 18 21 11 8 103 45 9 132 52 9 133 25 8 66 46 21 58 41 21 51 37 21 72 7

catalysts are collected in Table 1, together with those of the reference undoped materials (T_0 series). 3.1.1. XRPD Characterization. XRPD analysis clearly showed that doping or codoping TiO2 with fluorine inhibits the anatase into rutile phase transition, even for very low dopant amounts (see the F_3 series). In fact, all samples doped or codoped with fluorine, even if calcined at 700 °C, consisted of almost pure anatase phase. This confirms the crystal phase controlling role played by fluorine doping, resulting in a retarded condensation of spiral chains of rutile TiO6 octahedra at calcination temperatures above 500 °C.16−18 It is worth underlining that this retardation effect does not occur in the case of the singly N-doped material, the amount of residual anatase phase in N_5_700 being only 19% (Table 1), i.e., even lower than that in undoped T_0_700. Different results were obtained in the case of doping TiO2 with H3BO3. First, the brookite phase was detected in B-doped samples calcined at 500 °C, especially for relatively high nominal dopant amounts (B_25_500). Second, full anatase materials were obtained after calcination at 700 °C only by employing a relatively large nominal amount of H3BO3, i.e., in the B_25 photocatalyst series. In the case of the BF-codoped systems, full anatase materials were generally obtained, but

The suspensions were stirred in the dark for 15 min, to attain the adsorption equilibrium of the substrate on the photocatalyst surface, and then irradiated for 80 min under stirring. The irradiation source was a 300 W xenon lamp (Hamamatsu Photonics C2578-02) within a diffraction grating-type illuminator (Jasco CRM-FD). The irradiation wavelength was selected in the 370−460 nm range, with a full-width at halfmaximum intensity (fwhm) smaller than ca. 17 nm. The irradiation intensity, measured by a Hioki 3664 optical power meter, was in the 15−20 mW range. All other conditions were similar to those of photocatalytic activity tests under highpressure mercury lamp irradiation. The wavelength-dependent apparent quantum efficiency Φapp was calculated as the ratio between the rate of photogenerated holes consumption and the flux of incident photons, by taking into account that eight electrons (holes) are required for full oxidation of an acetic acid molecule into two molecules of CO2 and H2O.

3. RESULTS AND DISCUSSION 3.1. Photocatalyst Structure. The main structural features, determined by XRPD, BET, and XPS analyses, of the investigated series of doped and codoped TiO2 photo25581

dx.doi.org/10.1021/jp5084696 | J. Phys. Chem. C 2014, 118, 25579−25589

The Journal of Physical Chemistry C

Article

The absorption spectra of TiO2 singly doped with boron or codoped with fluorine and boron do not show any significant extra absorption in the visible region, as shown in Figure 2 for

traces (around 2−5%) of rutile phase appeared after calcination at 700 °C. The dimensions of anatase crystallites, calculated according to the Scherrer equation, did not sensibly depend on the nominal amount of fluorine in the F-doped series, whereas Band BF-doping seems to favor the formation of slightly smaller anatase TiO2 particles. In general, crystallite dimensions increased with increasing calcination temperature (Table 1). An unusual trend of anatase crystallite dimensions decreasing with increasing the nominal dopant content, in particular, in samples calcined at 700 °C, can be clearly observed in the case of BF-codoped samples. XRPD analysis of singly boron-doped or boron and fluorine codoped systems did not show any extra signal possibly associated with the presence of crystalline B2O3.19 3.1.2. Specific Surface Area. BET analysis showed a decrease in surface area of the TiO2-based photocatalysts with increasing the calcination temperature for a fixed nominal dopant amount, consequent to particles sintering. The SSA values of TiO2 samples containing fluorine (F and BF series) were similar to those of the previously investigated NFcodoped materials, for the same fluorine content and calcination temperature.6 B-doping appears beneficial for obtaining materials with higher SSA, if calcined at relatively low temperature (500 °C), in agreement with previous findings of lower crystallite dimensions and, consequently, higher SSA when boron is present in TiO2 materials.20 3.1.3. DR Characterization. A comparison between the absorption spectra of the F-doped TiO2 materials with those of the previously investigated NH4F-doped ones (NF-doped TiO2, labeled as D_TiO2 in ref 6) evidences that the absorption contribution with a maximum around 420 nm (labeled B in Figure 1) does not appear in the absorption spectra of the F-

Figure 2. Absorption spectra of (a) B_5 and (b) BF_5 doped TiO2 series calcined at different temperatures.

the B_5 and BF_5 series. When calcined at 700 °C, such materials exhibit a slightly red-shifted absorption edge due to the presence of the rutile phase, which is more marked in the case of B_5_700 (Figure 2), in full agreement with its higher rutile content detected by XRPD analysis (see Table 1). The presence of the fluorine dopant, also together with boron as codopant, does not produce TiO2-based materials showing extra visible light absorption, in line with literature reports.17,21−23 3.1.4. XPS Characterization. The presence and oxidation state of the fluorine and boron dopants and their main effects on the surface properties of TiO2 were investigated by XPS analysis. In particular, we focused on the undoped material in comparison with the F_12 and BF_12 samples calcined at 500 and 700 °C, and on B_12_600 and B_25_600, among the only B-doped TiO2 materials. The O/Ti and dopant/Ti surface atomic ratios, determined by quantitative XPS analysis, are collected in Table 2. First of all, almost identical Ti 2p doublet signals, with the two components at BEs of 464.7 and 459.0 eV, assigned to Ti 2p1/2 and Ti 2p3/2, respectively, were recorded with both the undoped and the differently doped samples. On the contrary, peculiar differences were observed in the O 1s spectra (Figure 3). In the case of pure TiO2 and F-doped TiO2 (see the XPS data of T_0_500 and F_12_500 in Figure 3), the main peak at 530.2 eV, typical of oxygen in the TiO2 lattice, and a shoulder at higher BE (531.6 eV), which originated from surface hydroxyl groups,24 were observed. In the case of boron-containing materials and especially for the BF_12 photocatalysts, a third O 1s component, centered at 532.8 eV, also appeared (Figure 3). This was accompanied by a variation of the O/Ti ratio, which increased from a nearly stoichiometric value of 2 obtained for undoped and F-doped TiO2 to a value even higher than 3, as reported in Table 2. Moreover, the relative percent

Figure 1. Comparison between the absorption spectrum of NF-doped TiO2 and the corresponding singly F-doped sample, both containing 25 mol % of dopant and calcined at 500 °C.

doped samples. This confirms that such an absorption feature, which was evidenced by spectral difference analysis for NFcodoped materials calcined at 500 °C, but not for those calcined at 700 °C, and found to be inactive in acetic acid degradation, is originated by nitrogen doping.6 On the other hand, a second absorption contribution, labeled A in Figure 1, with a maximum around 365 nm, was evidenced by spectral subtraction analysis in the near-UV region of both F-TiO2 and NF-codoped TiO2 series. This absorption contribution increases with increasing the calcination temperature of the materials and appears to be related to structure effects induced by the presence of the fluorine dopant. 25582

dx.doi.org/10.1021/jp5084696 | J. Phys. Chem. C 2014, 118, 25579−25589

The Journal of Physical Chemistry C

Article

Table 2. XPS Analysis of Selected Photocatalysts, in Terms of Surface O/Ti and Dopant/Ti Atomic Ratios O/Ti ratio

F/Ti ratio (%)

B/Ti ratio (%)

sample

as prepared

after sputtering

as prepared

after sputtering

T_0_600 T_0_700 F_12_500 F_12_700 BF_12_500 BF_12_700 B_12_600 B_25_600

2.3 2.4 2.3 2.4 3.5 2.4 2.4 2.7

1.6 1.8 3.3 2.2 2.1 2.3

1.7 0.9 0.5 0.1

0.5 0.1 0.2

as prepared

after sputtering

13.2 8.5 5.9 11.8

3.1 1.3 4.3 5.7

A confirmation of this hypothesis was obtained from the analysis of the B 1s XPS data recorded with the BF-codoped materials before and after etching with Ar+ ions (see Figure 4).

Figure 3. XPS spectra in the O 1s region of pure TiO2 (T_0_500) or of samples containing different nominal dopant/Ti percent molar ratios.

Figure 4. XPS spectra in the B 1s region of B-doped and BF-codoped TiO2 samples before (solid line) and after (dotted lines) etching with Ar ions.

contribution of this component, obtained by applying a deconvolution procedure to the O 1s XPS signal shown in Figure 3 (see Table 3), increased with increasing the B content (Table 2). Therefore, the O 1s XPS component located at the highest BE can be unambiguously assigned to the presence of B−O bonds25 and reveals the possible formation of B2O3 microaggregates at the TiO2 surface.26 Surface hydration (shoulder at 531.6 eV) appears to increase in parallel.

Quite intense XPS signals were recorded in this BE region before etching, the B/Ti atomic ratio being 13.2 and 8.5 for BF_12_500 and BF_12_700, respectively; i.e., the sample calcined at 500 °C had a surface boron content higher than the nominal one. In both cases, the B 1s signal consisted of a single peak centered at 192.8 eV, whose position and fwhm are in agreement with the formation of full oxidized boron

Table 3. Deconvolution of O 1s XPS Signals Recorded with Selected Samples, Either As Prepared or after Washing sample F_0_500 F_12_500 BF_12_500 B_12_600 B_25_600 BF_12_500 washed B_25_600 washed

532.8 eV (B−O) (%)

531.6 eV (surface OH) (%)

530.2 eV (Ti−O) (%)

17 12 15

9 12 22 13 17 12 14

91 88 61 75 68 88 84

2 25583

dx.doi.org/10.1021/jp5084696 | J. Phys. Chem. C 2014, 118, 25579−25589

The Journal of Physical Chemistry C

Article

Figure 5. XPS spectra in the F 1s region of BF_12_500 and F_12_500.

Figure 6. XPS spectra in the (a) B 1s and (b) F 1s regions of BF_12_500 before (gray line) and after washing (black line).

(B2O3).20,24−28 The absence of evident peak asymmetry is incompatible with the presence of either interstitial or substitutional B atoms in the TiO2 lattice, in agreement with DFT calculations29 and recent experimental results showing the instability of such B dopant sites on an oxygen-rich surface,26 especially after thermal treatment. The same conclusion can be drawn for the singly B-doped samples, showing similar B 1s XPS signals as the BF-codoped ones (Figure 4). However, for a fixed amount of B dopant, the copresence of fluorine in the doped material appears to facilitate surface boron segregation, in the B2O3 form. In fact, both BF_12 samples showed higher surface boron content with respect to the B_12_600 photocatalyst, as shown in Table 2. Furthermore, the amount of boron detected after sputtering in samples singly doped with boron was higher than that detected after sputtering in the BF-codoped samples. In particular, upon sputtering with Ar ions, a 27% decrease of the boron content is observed for B_12_600, to be compared with an above 75% decrease for both BF_12 samples. Therefore, fully oxidized boron species persist in the deeper TiO2 layers only in B-doped materials (B_12_600 and B_25_600).

The F 1s XPS data of BF_12_500 and F_12_500 are shown in Figure 5. The F 1s XPS peak consisted of a single, rather asymmetric band, at a BE of ca. 684 eV, attributed to F− ions adsorbed on the photocatalyst surface,16,21,23 which decreased after etching with Ar+ ions. The signal deconvolution shows that the main peak asymmetry is due to the presence of a minor component at 685.5 eV (Figure 5), which was already observed on F-doped TiO2 and assigned to fluorine atoms that replaced two −OH groups coordinated to the same Ti atom. An XPS signal at 688.5 eV, assigned to substitutional F ions in the TiO2 lattice,16,17 appeared only in the case of BF_12_500. Such a signal could not be detected in other samples containing fluorine, either before or after etching, probably being below the detection limit of the XPS technique. The F/Ti ratio, determined by quantitative XPS analysis for both F_12 and BF_12 samples (Table 2), was much lower than their nominal content (12%), especially in the case of fluorine and boron codoped TiO2. A comparison between the F 1s XPS signal detected for F_12 and BF_12 samples evidences that the amount of fluoride ions on the TiO2 surface decreased with increasing the 25584

dx.doi.org/10.1021/jp5084696 | J. Phys. Chem. C 2014, 118, 25579−25589

The Journal of Physical Chemistry C

Article

undoped TiO2 (T_0 series) are also shown in Figures 7 and 8, for comparison.

calcination temperature (Table 2). Furthermore, the F 1s XPS peak of F_12_700 and BF_12_700 almost completely disappeared after etching. These results are in line with those previously obtained for NF-doped samples6 and exclude the possibility of detecting fluorine in bulk F- or BF-TiO2, especially in samples calcined at high temperature. The two photocatalyst samples BF_12_500 and B_25_600, exhibiting the most remarkable shoulder in the O 1s XPS spectra (see Figure 3), underwent an acidic washing treatment, aimed at removing surface B2O3 microaggregates. As a matter of fact, the signal in the B 1s BE region completely disappeared from the XPS spectra of both washed samples (see, for example, Figure 6a), with a simultaneous restoring of the stoichiometric O/Ti ratio, accompanied by the loss of the O 1s signal component centered at 532.8 eV (Table 3). At the same time, in the case of BF_12_500, the disappearance of the B 1s signal was accompanied by the appearance of an intense F 1s XPS signal located at about 684 eV, thus attributed to terminal Ti−F bonds (Figure 6b). Furthermore, an effective F content higher than the nominal one was detected, the F/Ti ratio being 0.166. Thus, the washing procedure demonstrated a peculiar surface segregation of F− ions just below the B2O3 layer in fluorine and boron codoped materials. 3.2. Photoactivity Tests under Polychromatic Irradiation. Formic acid is an excellent degradation substrate to test photoactivity because (i) it does not absorb in the visible region, thus allowing a straightforward evaluation of the photocatalysts’ visible light activity, and (ii) it undergoes direct photomineralization without forming any stable intermediate species, which simplifies the interpretation of kinetic results. The photocatalytic degradation of this test compound occurred at a constant rate, i.e., according to a zero-order rate law, as in previous studies.5−8,15,30 Therefore, the activity in formic acid photocatalytic oxidation of the here investigated doped and codoped TiO2-based photocatalysts can be compared in Figure 7, in terms of zero-order rate constants, k0.

Figure 8. Zero-order rate constants of CO2 photoevolution during acetic acid decomposition on photocatalysts containing different amounts/types of dopants and calcined at different temperatures.

First of all, with moderately doped TiO2, both reactions confirmed to proceed at a higher rate with respect to that attained with undoped TiO2.5−8 Moreover, with the undoped and with the B_5 photocatalyst series, the reaction rate decreased with increasing the calcination temperature, most probably as a consequence of the transformation of the anatase into the less active rutile phase (Table 1). By contrast, the rate of both test reactions increased with increasing the calcination temperature of doped or codoped materials, especially in the case of the F series. 3.2.1. Effects of Doping with Fluorine. The results obtained with the F-doped and BF-codoped photocatalysts are perfectly in line with those obtained with NF-doped TiO2.5−7 This remarkable similarity suggests that the beneficial effects on photoactivity obtained with doped TiO2 samples calcined at high temperature is mainly related to structural modifications induced by fluorine, rather than by the presence of nitrogen or boron as codopants. This is better outlined in Figure 9A, where the rate constants of acetic acid decomposition obtained with TiO2-based photocatalysts containing a fixed (5 mol %) nominal amount of each single dopant source and calcined at 500 and at 700 °C are compared. The rate constants obtained when employing the photocatalysts calcined at 500 °C, all consisting of almost pure anatase, are quite similar, falling in the (0.45 ± 0.07) μmol CO2 min−1 range, if B-TiO2 is considered an outlier. On the other hand, the rate constants in acetic acid decomposition obtained with photocatalyst samples calcined at 700 °C are remarkably different. Not only are the values obtained with materials not containing fluorine much lower than those of materials doped or codoped with fluorine, but they are also lower than the rate constant values obtained with the same materials calcined at 500 °C. NF- and BF-codoped materials evidence similar photoactivity, significantly lower than that of singly fluorinedoped F_5_700. Better inspection in the data shown in Figure 9A reveals that the rate constants increase with the SSA of the materials. In fact, the photoactivity, in terms of k0 rate constants in acetic acid decomposition, of the F-, N-, and B-doped or codoped materials calcined at 500 °C, all consisting of almost pure anatase phase (Table 1), exhibit a linear correlation with their SSA, as shown in Figure 9B. Thus, the beneficial effect in

Figure 7. Zero-order rate constants of formic acid photomineralization on photocatalysts containing different amounts/types of dopants and calcined at different temperatures.

The oxidative decomposition of transparent acetic acid was also investigated to check the photoactivity scale of the materials. In this case, the photocatalytic evolution of CO2 from suspensions containing acetic acid and either pure or doped TiO2 samples always occurred at a constant rate, and the zeroorder rate constants k0 obtained from these photocatalytic activity tests are collected in Figure 8. Results obtained with 25585

dx.doi.org/10.1021/jp5084696 | J. Phys. Chem. C 2014, 118, 25579−25589

The Journal of Physical Chemistry C

Article

calcined at 500 °C, but exhibit higher activity if they are fluorine-doped or codoped. It is worth underlining that all materials calcined at 500 °C consist of almost pure anatase, while the less performing materials calcined at 700 °C are the undoped (T_0_700) and the singly N- or B-doped (N_5_700 and B_5_700) materials, all containing significant amounts of rutile (Table 1). Only fluorine, among the here investigated p-block element TiO2 dopants, has an inhibition effect in the anatase into rutile phase transition. Indeed, both B_5_700 and N_5_700 contain significant amounts of rutile (36% and 81%, respectively, this second rutile content being even higher than that of T_0_700, the undoped TiO2 sample calcined at 700 °C), and all of these samples not containing fluorine exhibit very similar, low photoactivity and relatively low SSA. Furthermore, only fluorine-containing materials experience a photoactivity increase upon increasing the calcination temperature from 500 to 700 °C, with a maximum photoactivity increase in the case of singly fluorine-doped TiO2 (Figure 9B). This effect might be related to the hypothesized beneficial formation of surface oxygen defects under calcination at high temperature in the presence of the fluorine dopant,21,22 with the lower photoactivity increase of NF- and BF-codoped materials to be possibly related to a lower concentration of surface defects due to charge compensation effects in the presence of the two codopants.31 However, in the absence of any firm evidence for the formation of such surface defects, the photoactivity increase observed upon increasing the calcination temperature may be simply a consequence of the higher crystallinity of the materials calcined at 700 °C, which limits the recombination of the charge couples photoproduced in these materials. 3.2.2. Effects of Single Boron Doping. According to the photoactivity results shown in Figures 7 and 8, the use of only boron as dopant element of TiO2 does not appear to produce a systematic beneficial effect on the rates of both photocatalytic reactions. In particular, only B-TiO2 materials obtained by employing huge nominal dopant amounts (at least 12 mol %) and calcined at high temperature (700 °C) showed good photocatalytic activities. On the other hand, photocatalysts containing 5 mol % of boron showed an activity decrease with increasing the calcination temperature, as a consequence of both the remarkable decrease in SSA and the anatase into rutile phase transformation (Table 1).

Figure 9. (A) Zero-order rate constants k0 of CO2 photoevolution obtained during acetic acid decomposition on photocatalysts containing 5 mol % of different dopants and calcined at 500 or at 700 °C. (B) Correlation between k0 and the specific surface area of the doped or codoped TiO2 materials.

photoactivity produced by the presence of the different dopants appears to essentially derive from an increased surface area of the doped materials with respect to undoped TiO2, and no other effect, e.g., related to the electronic structure of doped materials, needs to be invoked. Furthermore, the presence of nitrogen and boron in NF- and BF-codoped materials calcined at 500 °C appears to have scarce effect on photoactivity, even in relation to their possible role as charge trap sites. Figure 9B shows that also the photoactivity of different dopant-containing samples calcined at 700 °C increases with their SSA, and with a much higher slope. These samples are all characterized by a lower SSA with respect to the same materials

Figure 10. Concentration vs irradiation time during the photocatalytic degradation of formic acid with BF_12_500 and B_25_600 photocatalysts, either as prepared (full symbols) or after the washing treatment (void symbols). 25586

dx.doi.org/10.1021/jp5084696 | J. Phys. Chem. C 2014, 118, 25579−25589

The Journal of Physical Chemistry C

Article

experimental conditions employed in the two test reactions. In fact, while acetic acid photooxidation was performed under UV light irradiation, the photodecomposition of formic acid was investigated employing an irradiation source mainly emitting visible light. Thus, a higher photoefficiency in the visible region might be evidenced, e.g., for the BF_5_700 photocatalyst. Therefore, aiming at verifying the possible activation in the visible region, the photooxidation of transparent acetic acid was systematically investigated as a function of irradiation wavelength in the 370−460 nm range, carrying out a so-called action spectra analysis. As in previous studies,6 we restricted this kind of analysis to full anatase TiO2 materials, in order to avoid any artifact due to the presence of rutile, this phase being less photoactive than anatase in acetic acid decomposition.36,37 The action spectra in the 370−460 nm range obtained with a 10 nm wavelength step for the F_5, BF_5, and B_25 doped TiO2 series are shown in Figure 11. None of the photocatalysts

Controversial results can be found in the literature concerning the origin of the improved photoactivity of Bdoped TiO2, to be partly related to different structural features of the materials. For instance, while Chen et al. observed a band-gap increase due to B-doping and attributed it to quantum-size effects,19 Zhao et al. detected a red-shift in the absorption spectrum of B-doped TiO2.32 This apparent contradiction can be explained considering the different geometry and electronic structure of B-doped TiO2, as recently shown by Geng et al.32 In fact, boron can either substitute oxygen, or sit in interstitial positions, or also substitute Ti4+ in the lattice. DFT calculations indicate that the last case is less favorable, while the other two have comparable energy and both of them may occur.33 Notably, only B- for O-substitution will lead to band-gap narrowing, while the occupation of an interstitial site will produce a blue-shift of the absorption properties of the doped materials.32,33 Furthermore, TiO2-doping with boron was found to enhance the hydration ability of the materials,34 which may strongly affect the adsorption capability of substrate molecules and, consequently, the overall photocatalytic reaction efficiency. At the same time, the possible formation of surface B2O3, characterized by a 5.02 eV band gap, was also invoked to explain the photoactivity increase attained with boron-doping.35 In fact, the so-obtained TiO2/B2O3 interface may act as a separation site for photogenerated electron−hole couples, thus suppressing their undesired recombination.19,21,35 In order to get a better insight into the possible role of surface B2O3 species on TiO2 photoactivity, we compared the photoactivity of our as prepared B-doped TiO2 with that of the same materials after washing with an acidic solution. This treatment produced the complete removal of B2O3, according to XPS analysis. Examples of the kinetic results obtained with B_25_600 and BF_12_500 in formic acid degradation are shown in Figure 10. The concentration vs irradiation time profiles show that surface B2O3, formed upon TiO2 doping with boron, does not have any beneficial or detrimental effect on the photoactivity of both B- or BF-doped TiO2, complete formic acid degradation being attained after the same irradiation time for both washed and unwashed photocatalysts (Figure 10). On the other hand, with the washed materials, the reaction started with a higher rate, especially in the case of BF_12_500. A possible explanation of the short initial induction period observed with the unwashed materials is that surface B2O3 may undergo dissolution under the acidic conditions of formic acid degradation. This points out serious doubts about the real beneficial effects of surface B2O3 on the photocatalytic efficiency of TiO2. On the other hand, XPS analysis indicates that full oxidized boron species may persist within B-doped TiO2 (Figure 4). The unclear photoactivity trend with both the nominal dopant content and calcination temperature observed in the two test reactions for the B_12 and B_25 photocatalysts (Figures 7 and 8) may simply result from the counterbalancing effects of Binduced anatase phase stabilization and surface area decrease with increasing the calcination temperature (Table 1). 3.3. Action Spectra Analysis of Acetic Acid Decomposition in the 370−460 nm Range. The general trend of photocatalytic activity increase with increasing the calcination temperature for a fixed nominal dopant amount is more evident in formic acid degradation (Figure 7) than in acetic acid decomposition (Figure 8). This could be related to the different

Figure 11. Action spectra of acetic acid decomposition in the 370−460 nm wavelength range of the (a) F_5, (b) BF_5, and (c) B_25 doped TiO2 photocatalysts series.

exhibit photoactivity in the visible light region, i.e., at λ > 420 nm, as expected from the absence of any absorption component under visible light induced by either fluorine and/or boron doping (Figures 1 and 2). On the other hand, Figure 11a,b clearly shows that, for only the F_5 and BF_5 photocatalyst series, a progressively higher calcination temperature ensured a higher apparent quantum efficiency in the near UVA region, with a trend almost identical to that observed in the case of NFcodoped TiO2.6 On the contrary, this kind of beneficial effect was not found in the case of single boron doping (Figure 11c). 25587

dx.doi.org/10.1021/jp5084696 | J. Phys. Chem. C 2014, 118, 25579−25589

The Journal of Physical Chemistry C

Article

(4) Dozzi, M. V.; Selli, E. Doping TiO2 with p-block elements: Effects on photoactivity. J. Photochem. Photobiol., C 2013, 14, 13−28. (5) Dozzi, M. V.; Ohtani, B.; Selli, E. Absorption and action spectra analysis of ammonium fluoride-doped titania photocatalysts. Phys. Chem. Chem. Phys. 2011, 13, 18217−18227. (6) Dozzi, M. V.; Livraghi, S.; Giamello, E.; Selli, E. Photocatalytic activity of S- and F-doped TiO2 in formic acid mineralization. Photochem. Photobiol. Sci. 2011, 10, 343−349. (7) Dozzi, M. V.; Saccomanni, A.; Altomare, A.; Selli, E. Photocatalytic activity of NH4F-doped TiO2 modified by noble metal nanoparticle deposition. Photochem. Photobiol. Sci. 2013, 12, 595−601. (8) Dozzi, M. V.; D’Andrea, C.; Ohtani, B.; Valentini, G.; Selli, E. Fluorine-doped TiO2 materials: Photocatalytic activity vs timeresolved photoluminescence. J. Phys. Chem. C 2013, 117, 25586− 25595. (9) Reyes-Garcia, E. A.; Sun, Y.; Raftery, D. Solid-state characterization of the nuclear and electronic environments in a boron-fluoride co-doped TiO2 visible-light photocatalyst. J. Phys. Chem. C 2007, 111, 17146−17154. (10) Fittipaldi, M.; Gombac, V.; Gasparotto, A.; Deiana, C.; Adami, G.; Barreca, D.; Montini, T.; Martra, G.; Gatteschi, D.; Fornasiero, P. Synergistic role of B and F dopants in promoting the photocatalytic activity of rutile TiO2. ChemPhysChem 2011, 12, 2221−2224. (11) Li, F.; Wang, X.; Zhao, Y.; Liu, J.; Hao, Y.; Liu, R.; Zhao, D. Ionic-liquid-assisted synthesis of high-visible-light-activated N−B−Ftri-doped mesoporous TiO2 via a microwave route. Appl. Catal., B 2014, 144, 442−453. (12) Ohtani, B. Preparing articles on photocatalysis − Beyond the illusions, misconceptions, and speculation. Chem. Lett. 2008, 37, 217− 229. (13) Ohtani, B. Photocatalysis A to Z − What we know and what we do not know in a scientific sense. J. Photochem. Photobiol., C 2010, 11, 157−178. (14) Rietveld, H. M. Profile refinement method for nuclear and magnetic structures. J. Appl. Crystallogr. 1969, 2, 65−71. (15) Dozzi, M. V.; Prati, L.; Canton, P.; Selli, E. Effects of gold nanoparticles deposition on the photocatalytic activity of titanium dioxide under visible light. Phys. Chem. Chem. Phys. 2009, 11, 7171− 7180. (16) Yu, J. C.; Yu, J.; Ho, W.; Jiang, Z.; Zhang, L. Effects of F-doping on the photocatalytic activity and microstructures of nanocrystalline TiO2 powders. Chem. Mater. 2002, 14, 3808−3816. (17) Li, D.; Ohashi, N.; Hishita, S.; Kolodiazhnyi, T.; Haneda, H. Origin of visible-light-driven photocatalysis: A comparative study on N/F-doped and N-F-codoped TiO 2 powders by means of experimental characterizations and theoretical calculations. J. Solid State Chem. 2005, 178, 3293−3302. (18) Dozzi, M. V.; Selli, E. Specific facets-dominated anatase TiO2: Fluorine-mediated synthesis and photoactivity. Catalysts 2013, 3, 455− 485. (19) Chen, D.; Yang, D.; Wang, Q.; Jiang, Z. Effects of boron doping on photocatalytic activity and microstructure of titanium dioxide nanoparticles. Ind. Eng. Chem. Res. 2006, 45, 4110−4116. (20) Gombac, V.; De Rogatis, L.; Gasparotto, A.; Vicario, G.; Montini, T.; Barreca, D.; Balducci, G.; Fornasiero, P.; Tondello, E.; Graziani, M. TiO2 nanopowders doped with boron and nitrogen for photocatalytic applications. Chem. Phys. 2007, 339, 111−123. (21) Li, D.; Haneda, H.; Labhsetwar, N. K.; Hishita, S.; Ohashi, N. Visible-light-driven photocatalysis on fluorine-doped TiO2 powders by the creation of surface oxygen vacancies. Chem. Phys. Lett. 2005, 401, 579−584. (22) Ho, W.; Yu, J. C.; Lee, S. Synthesis of hierarchical nanoporous F-doped TiO2 spheres with visible light photocatalytic activity. Chem. Commun. 2006, 1115−1117. (23) Czoska, A. M.; Livraghi, S.; Chiesa, M.; Giamello, E.; Agnoli, S.; Granozzi, G.; Finazzi, E.; Di Valentin, C.; Pacchioni, G. The nature of defects in fluorine-doped TiO2. J. Phys. Chem. C 2008, 112, 8951− 8956.

The higher apparent quantum efficiency in acetic acid decomposition under UVA light parallels the increase of the absorption contribution labeled A in Figure 1, which increased with increasing calcination temperature. The observed trend of photoactivity increase, mainly centered in the 370−400 nm range (Figure 11a,b) thus appears to be peculiar of all F-doped or codoped TiO2 photocatalysts; i.e., it is a consequence of TiO2 doping with fluorine (not with other codoped p-block elements). Fluorine favors the formation of a stable, highly crystalline anatase phase after calcination at high temperature, ensuring long-living photoproduced charge couples, as recently demonstrated by time-resolved photoluminescence experiments.8

4. CONCLUSIONS The following conclusions can be drawn from the present study: (i) TiO2 moderately doped (5−12 mol %) with N, F, and B or codoped with N and F or with B and F, prepared by a sol−gel synthesis and calcined at relatively low temperature (500 °C), exhibit increased photocatalytic activity with respect to undoped TiO2 essentially due to their lower particle size and increased specific surface area. The electronic structure of these doped materials does not need to be invoked to account for their photoactivity increase with respect to undoped TiO2. (ii) Samples calcined at higher temperature (700 °C) exhibit increased photoactivity in the near UVA region only if they contain fluorine as dopant or codopant. Such materials mainly consist of highly crystalline anatase, ensuring a lower rate of photoproduced electron−hole recombination. (iii) Surface B2O3 formed upon TiO2 doping with boron has no effect on photoactivity. The observed photoactivity increase upon Bdoping TiO2 may simply result from the increased surface area and retarded anatase into rutile transformation of the doped materials with respect to undoped TiO2.



AUTHOR INFORMATION

Corresponding Author

*Phone: +39 02 50314237. Fax: +39 02 50314300. E-mail: [email protected] (E.S.). Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS Financial support from the Cariplo Foundation through the project Novel photocatalytic materials based on heterojunctions for solar energy conversion is gratefully acknowledged. This work has been also funded by the Italian Ministry of Instruction, University and Research (MIUR) through the FIRB Project RBAP115AYN Oxides at the nanoscale: multif unctionality and applications.



REFERENCES

(1) Asahi, R.; Morikawa, T.; Ohwaki, T.; Aoki, K.; Taga, Y. Visiblelight photocatalysis in nitrogen-doped titanium oxides. Science 2001, 293, 269−271. (2) Khan, S. U. M.; Al-Shahry, M.; Ingler, W. B., Jr. Efficient photochemical water splitting by a chemically modified n-TiO2. Science 2002, 297, 2243−2245. (3) Kumar, S. G.; Devi, L. G. Review on modified TiO 2 photocatalysis under UV/visible light: selected results and related mechanisms on interfacial charge carrier transfer dynamics. J. Phys. Chem. A 2011, 115, 13211−13241. 25588

dx.doi.org/10.1021/jp5084696 | J. Phys. Chem. C 2014, 118, 25579−25589

The Journal of Physical Chemistry C

Article

(24) Artiglia, L.; Zana, A.; Rizzi, G. A.; Agnoli, S.; Bondino, F.; Magnano, E.; Cavaliere, E.; Gavioli, L.; Granozzi, G. Water adsorption on different TiO2 polymorphs grown as ultrathin films on Pt(111). J. Phys. Chem. C 2012, 116, 12532−12540. (25) Lu, N.; Quan, X.; Li, J. Y.; Chen, S.; Yu, H. T.; Chen, G. Fabrication of boron-doped TiO2 nanotube array electrode and investigation of its photoelectrochemical capability. J. Phys. Chem. C 2007, 111, 11836−11842. (26) Artiglia, L.; Lazzari, D.; Agnoli, S.; Rizzi, G. A.; Granozzi, G. Searching for the formation of Ti−B bonds in B-doped TiO2−rutile. J. Phys. Chem. C 2013, 117, 13163−13172. (27) Liu, G.; Zhao, Y. N.; Sun, C.; Li, F.; Lu, G. Q.; Cheng, H. M. Synergistic effects of B/N doping on the visible-light photocatalytic activity of mesoporous TiO2. Angew. Chem., Int. Ed. 2008, 47, 4516− 4520. (28) Zaleska, A.; Sobczak, J. W.; Grabowska, E.; Hupka, J. Preparation and photocatalytic activity of boron-modified TiO2 under UV and visible light. Appl. Catal., B 2008, 78, 92−100. (29) Jin, H.; Dai, Y.; Wie, W.; Huang, B. Density functional characterization of B doping at rutile TiO2 (110) surface. J. Phys. D: Appl. Phys. 2008, 41, 195411. (30) Dozzi, M. V.; Chiarello, G. L.; Selli, E. Effects of surface modification on the photocatalytic activity of TiO2. J. Adv. Oxid. Technol. 2010, 13, 305−312. (31) Di Valentin, C.; Finazzi, E.; Pacchioni, G.; Selloni, A.; Livraghi, S.; Czoska, A. M.; Paganini, M. C.; Giamello, E. Density functional theory and electron paramagnetic resonance study on the effect of N− F codoping of TiO2. Chem. Mater. 2008, 20, 3706−3714. (32) Zhao, W.; Ma, W.; Chen, C.; Zhao, J.; Shuai, Z. Efficient degradation of toxic organic pollutants with Ni2O3/TiO(2−x)Bx under visible irradiation. J. Am. Chem. Soc. 2004, 126, 4782−4783. (33) Geng, H.; Yin, S. W.; Yang, X.; Shuai, Z. G.; Liu, B. G. Geometric and electronic structures of the boron-doped photocatalyst TiO2. J. Phys.: Condens. Matter 2005, 18, 87−96. (34) Moon, S. C.; Mametsuka, H.; Tabata, S.; Suzuki, E. Photocatalytic production of hydrogen from water using TiO2 and B/TiO2. Catal. Today 2000, 58, 125−132. (35) Yuan, J.; Wang, E.; Chen, Y.; Yang, W.; Yao, J.; Cao, Y. Doping mode, band structure and photocatalytic mechanism of B−N-codoped TiO2. Appl. Surf. Sci. 2011, 257, 7335−7342. (36) Torimoto, T.; Nakamura, N.; Ikeda, S.; Ohtani, B. Discrimination of the active crystalline phases in anatase-rutile mixed titanium(IV) oxide photocatalysts through action spectrum analyses. Phys. Chem. Chem. Phys. 2002, 4, 5910−5914. (37) Prieto-Mahaney, O. O.; Murakami, N.; Abe, R.; Ohtani, B. Correlation between photocatalytic activities and structural and physical properties of titanium(IV) oxide powders. Chem. Lett. 2009, 238−239.

25589

dx.doi.org/10.1021/jp5084696 | J. Phys. Chem. C 2014, 118, 25579−25589