rutile phase composition influence on the photocatalytic performances of mesoporous TiO2 powders

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e9

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.elsevier.com/locate/he

Evaluation of the anatase/rutile phase composition influence on the photocatalytic performances of mesoporous TiO2 powders  b, Elisabetta Masolo a,*, Nina Senes a, Eva Pellicer b, Maria Dolors Baro a a a Stefano Enzo , Maria Itria Pilo , Gabriele Mulas , Sebastiano Garroni a,* a b

Department of Chemistry and Pharmacy, University di Sassari and INSTM, Via Vienna 2, I-07100 Sassari, Italy Universitat Aut onoma de Barcelona, Departament de Fı´sica, E-08193 Bellaterra, Spain

article info

abstract

Article history:

We investigated the influence of the microstructure, texture and the relative amount of

Received 18 February 2015

anatase and rutile phases on the photoelectrocatalytic H2 evolution over mesoporous TiO2

Received in revised form

(m-TiO2). This study was carried out by means of ex-situ powder X-ray diffraction (XRPD),

27 May 2015

N2 physisorption, transmission electron microscopy (TEM), and photoelectrolysis mea-

Accepted 28 May 2015

surements. The synthesis of nanocrystalline TiO2 powders with different anatase/rutile

Available online xxx

weight ratios was successfully approached by evaporation-induced self assembly (EISA) method simply varying the volume amount of TiCl4 and Ti(OBu)4 (acidebase pair) as metal

Keywords:

precursors. In particular, m-TiO2 with anatase/rutile weight ratios similar to commercial

Titania

TiO2 (P25) was obtained. The microstructure, texture and photocatalytic properties toward

Water splitting

hydrogen evolution reaction (HER) of the as-synthesized TiO2 were compared with com-

EISA

mercial P25. Interestingly, the m-TiO2 with a high surface area of 196 m2/g, uniform pore

Rietveld analysis

size and 87/13 anatase/rutile weight ratio, showed a significant increase in the photo-

Transmission electron microscopy

activity with respect to the powders richer in either anatase or rutile phases. Furthermore,

(TEM)

it was claimed that the crystallinity and specific surface area play a key role on the pho-

Electrochemical properties

tocatalytic activity of the synthesized m-TiO2 samples. Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved.

Introduction The transition towards a hydrogen-based economy requires the development of cost-effective and free-carbon processes to produce pure hydrogen on a large scale. Nowadays, renewable energies impact on the commercial production of hydrogen corresponds only to 5%, while fossil fuels contribute

to the largest part (95%), hence making the process environmentally inconvenient [1]. The hydrogen production from renewable energies is both a scientific and technological challenge. Photocatalytic water splitting to produce hydrogen represents one of the most promising approaches since its photoelectrochemical production was demonstrated by Fujishima and Honda [2]. Among the materials exploited in photocatalysis, titania

* Corresponding authors. Tel.: þ39 079 229524; fax: þ39 079 212069. E-mail addresses: [email protected] (E. Masolo), [email protected] (S. Garroni). http://dx.doi.org/10.1016/j.ijhydene.2015.05.180 0360-3199/Copyright © 2015, Hydrogen Energy Publications, LLC. Published by Elsevier Ltd. All rights reserved. Please cite this article in press as: Masolo E, et al., Evaluation of the anatase/rutile phase composition influence on the photocatalytic performances…, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.05.180

2

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e9

(TiO2) is still the most widely used. Although its wide band gap (~3.2 eV) limits its photocatalytic activity mostly within the ultraviolet region of the electromagnetic spectrum it is very stable under photocatalytic conditions with respect to the corrosion and oxidation processes; moreover it is nontoxic, cheap and easy to find in nature [1,3e9]. The TiO2 equilibrium phase diagram presents three crystalline structures: anatase, rutile and brookite [10]. It is widely accepted that anatase is the most photoactive phase [11]. However, the presence of both anatase and rutile has proved to increase the photocatalytic performances of mixed phase TiO2 nanotubes [12]; moreover, the mixture of anatase and rutile in commercial TiO2 nanoparticles (Degussa P25) makes this material more active than other ones made of pure anatase [13]. P25 is composed of an anatase/rutile ratio of about 3:1 and its microstructure is characterized by the phase junction between these two polymorphs [13,14]. Nevertheless, potential effects of such features are only partially explored and the relationship between the structure and the photocatalytic activity of TiO2 is not straightforward disclosed due to the many physicoechemical parameters (microstructure, morphology, surface area, surface defects, texture, etc) involved in this process [15]. Specifically, an extended surface area with a high adsorbent surface density contributes to increase the performances of the photocatalytic reaction. On the other hand, the surface of a semiconductor is typically characterized by several defects such as impurities, oxygen vacancies, step edges, etc., which can act as trap sites, thus favouring the electronehole recombination by the SRH model (Shockley-Read-Hall Model) [16]. The electronehole recombination at the surface represents an important limiting step and a catalyst with a large surface area and a high number of defects (poor crystallinity) can lead to low photoelectrocatalytic activities [16,17]. For the aforementioned reasons, in order to decrease the defects in the microstructure while keeping a significant photocatalytic activity, TiO2 with a high crystalline framework and large surface area is highly desirable [18e21]. As a matter of facts ordered mesoporous materials can offer both large surface areas and great mass diffusion in their network [22,23]. One of the most promising and low cost approach for the synthesis of mesoporous titania seems to be the evaporation induced self assembly (EISA) method [24e26], where, starting from a diluted solution of a surfactant and titanium precursors, the slow evaporation of the reaction solvent, usually ethanol, leads to the formation of a gel consisting of surfactant micelles surrounded by just formed TiO2. The template is then removed usually by calcination and leaves well formed pores in the crystalline titania matrix. These syntheses are very sensitive to several parameters like

the humidity of the environment, the gelation and calcination temperature and the precursors choice. Usually a low pH level is required to slow down the reaction kinetics and allows the micelles formation before the complete hydrolysiscondensation of the metal precursors (usually an alcoxide) to TiO2. A low pH value is usually obtained by adding HCl to the reaction mixture or, as an alternative, by using a pH adjustor like TiCl4 together with a titanium alcoxide in a socalled acid-base approach [18,27]. EISA is often applied for the synthesis of titania thin films [25,28,29] but only a few works can be found concerning the synthesis of mesoporous powders [18,25,30,31] that can be more flexible for practical use. In this study, we investigate the effect of the anatase to rutile relative amount in mesoporous TiO2 (m-TiO2) powders on the photoelectrocatalytic production of hydrogen. The mTiO2 powders were prepared by an EISA route by tuning the concentration of the TiCl4/Ti(OBu)4 acid-base pair, used as metal precursors. The as-obtained matrix was then calcined to remove the templating agent (P123) and the powders were investigated in detail by X-Ray Powder Diffraction (XRPD), N2 physisorption and Transmission Electron Microscopy (TEM). These techniques combined with the photocatalytic tests helped to evaluate and understand the photoactivity of mTiO2 in relation to the anatase and rutile microstructure.

Experimental details Commercial powders of TiO2 (purity > 99.5%), liquid TiCl4 and Ti(OBu)4 were purchased from SigmaeAldrich. In a typical synthesis, 1 g of surfactant was dissolved in 20 mL of ethanol and then 0.78 mL of Ti(OBu)4 and 2.0, 1.0 or 0.5 mL of TiCl4 were added to the solution, which was afterward kept under stirring for 2 h at room temperature [22]. The mixture was then transferred into a Petri dish and put in an oven at 40  C with a moisture level between 40 and 60 %, for 24 h to promote gelification. The obtained gel was finally calcined in a muffle at 350  C (1  C min1) for 6 h and then allowed to slowly cool down to room temperature. The relative amounts of the reagents in the starting mixtures are reported in Table 1. All the characterizations are done on synthesised titania powders after heat treatment at 400  C for 1 h. As a general procedure for the preparation of the electrode materials, 100 mg of each sample were grinded for 2 min together with 1.6 mL of ethanol and 60 mg of polyethylene glycol (PEG, 20,000 g/mol) and then the mixture was sonicated for 10 min to ensure a good dispersion of the powders. The suspension was then deposited into an Indium Tin Oxide (ITO) glass with a conductive area of 1.5 cm2 by the doctor blade

Table 1 e TiO2 Anatase/Rutile weight % composition, corresponding crystallite sizes and specific surface area. Commercial P25 data are also reported for comparison. TiCl4/Ti(OBu)4 (ml) P25 (commercial) 2.0/0.78 1.0/0.78 0.5/0.78

lonset (nm)

Band Gap (eV)

J (105 A cm2)

H2 evolution rate (mmol h1 cm2)

420 403 410 390

2.95 3.07 3.02 3.18

8.47 3.03 3.67 3.39

~10 3.19 4.87 3.20

Please cite this article in press as: Masolo E, et al., Evaluation of the anatase/rutile phase composition influence on the photocatalytic performances…, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.05.180

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e9

technique using two pieces of adhesive tape to control the thickness of the titania layer. The TiO2/ITO glass was subsequently thermally treated in a muffle at 400  C for 1 h and then allowed to cool to room temperature during 3 h. The structural characterization of all the samples was carried out by XRPD (Rigaku DMax diffractometer equipped with a graphite monochromator on the diffracted beam), and Transmission Electron Microscopy TEM (Jeol-JEM 2011, 200 kV). For TEM observations, samples were prepared by dispersing a small amount of powder in ethanol followed by the deposition of one or two drops of the suspension on a holey carbon supported grid. Phase abundance and microstructural parameters were evaluated by fitting the XRPD patterns by Rietveld method with the software MAUD [32]. N2 sorption isotherms were collected on a Sorptomatic 1990 instrument and the data were analyzed according to BET and BJH methods. Around 200 mg of sample were charged in a quartz tube and degassed under high vacuum (1  103 bar) at 250  C for 24 h. The empty volume in the quartz tube was evaluated by helium sorption isotherms. UVevis spectra were recorded on a water suspension of each sample (1 mg of titania in 100 mL) using a Hitachi U-2010 equipment and the band gap was evaluated from the lonset that arised from the analysis of the absorption spectra with the tangent method, and then by applying the Plank Equation. All (photo)electrochemical characterizations were performed in a home-made single-compartment three electrode quartz cell under inert atmosphere in a previously degassed NaOH 1M solution using the prepared TiO2/ITO as working electrodes, a platinum wire as auxiliary electrode and Ag/AgCl as reference electrode (E ¼ 207.6 mV vs NHE). Linear Sweep

3

Voltammetry (LSV) tests were recorded at a scan rate of 5 mV s1. Experiments were run starting from the open circuit potential of the system, first in the dark and then under direct illumination, and controlled by an Autolab PGSTAT 12 (EcoChemie) potentiostat/galvanostat interfaced with a PC under General Purpose Electrochemical Systems (GPES) software. Potentiostatic photoelectrolysis experiments were carried out under direct illumination for 1 h at a potential value chosen from the LSV scan. The amount of H2 considering the reduced area (<1.5 cm2) of the ITO glass where the m-TiO2 was deposited and the low photoactivity under sun illumination for non-doped titania, is theoretically obtained according to the Faraday's Law [33], considering the charge (in Coulomb) flowed during the electrolysis at 0.9 V vs Ag/AgCl and the visible H2 evolution from the water splitting reaction. The light source was provided by an ABET TECHNOLOGIES Sun 2000 Solar Simulator with a irradiating power of 1 sun (AM 1.5).

Results and discussion The XRPD patterns of the TiO2 samples prepared by varying the acid-base TiCl4eTi(OBu)4 metal precursor volumes are reported in Fig. 1, together with the profile relevant to commercial P25 subjected to the heat treatment at 400  C (HT 400  C), here displayed for comparison. In order to estimate the microstructural parameters of the phases contributing to the patterns with a high degree of accuracy, the full diffractograms (Fig. 1A) were carefully analyzed. It can be seen that the Rietveld profiles (pink and red lines) interpolate the experimental data points (blue dots)

Fig. 1 e A. CuKa X-ray diffraction patterns of the samples prepared with variable volume ratios (mL) of TiCl4eTi(OBu)4. B. Zoom in the angular 2q range 20e60 , which gives an idea of the accuracy of the Rietveld refinement. Blue dots, pink, red and light green lines correspond to the experimental pattern, anatase and rutile phases contribution, and the pattern baseline, respectively. The pattern of the HT 400  C commercial P25 is shown for comparison. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) Please cite this article in press as: Masolo E, et al., Evaluation of the anatase/rutile phase composition influence on the photocatalytic performances…, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.05.180

4

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e9

with a high degree of accuracy. The XRPD analyses reveal that the powders consist of a mixture of anatase (A) and rutile (R) phases. As shown in detail in Fig. 1B, the Bragg peaks of the ATiO2 phase simultaneously increase in intensity with respect to the rutile peaks, as the TiCl4/Ti(OBu)4 ratio decreases. The quantitative data gleaned from the Rietveld treatment of the XRPD patterns are summarized in Table 1, where the phase weight % abundance and crystallite sizes are reported for each phase in the samples. In particular, the pattern corresponding to the TiO2 synthesized with a 1.0/0.78 mL TiCl4/Ti(OBu)4 ratio shows an A:R weight % of 87:13, similar to that of commercial P25 (83:17, see Table 1). Moreover, it is possible to appreciate a clear relationship between the amount of TiCl4 used in the synthesis and the corresponding A:R ratio: keeping the Ti(OBu)4 volume constant (0.78 mL), the anatase content decreases (74:26 A:R) or increases (92:8 A:R) using, respectively, higher (2.0 mL) or lower (0.5 mL) TiCl4 volumes as reported in Table 1, and this is totally in agreement with the previous work published on this system [22]. Moreover, the shape of XRPD peaks shown in Fig. 1B, together with the results arising from the Rietveld refinement,

suggests that the microstructural parameters of both the anatase and rutile phases are quite similar and do not vary upon changing the TiCl4/Ti(OBu)4 volume ratios. Contrarily, the rutile shows larger crystallite size (R: 525 Å) than the ˚ ) in the P25 sample (see Table 1). anatase counterpart (A: 350 A It is worth mentioning that the previously cited characteristics allow us to study the influence of the anatase/rutile weight ratios on the photocatalytic activity of the assynthesized TiO2, excluding any eventual contribution arising from differences in the crystallite size. Representative TEM images of the commercial P25 and the synthesized TiO2 samples with different anatase and rutile amounts are presented in Fig. 2. The commercial P25 (Fig. 2A) shows homogenous particle sizes ranging from 20 to 40 nm, characterized by a rhombic- and square-like shapes, in accordance to the data reported by Martra et al. [30]. The rhombic-like shape can be associated with a truncated bipyramid particle with the longer axis oriented perpendicular to the electron beam (see inset in Fig. 2A). The size distribution of the particles is not influenced by the presence of mixed phases as it is uniform when referring

Fig. 2 e TEM images of the commercial P25 and TiO2 samples prepared from different relative amounts of TiCl4 and Ti(OBu)4 (in mL): A. P25. B. 2.0/0.78 TiCl4/Ti(OBu)4 ratio. C. 1.0/0.78 TiCl4/Ti(OBu)4 ratio. D. 0.5/0.78 TiCl4/Ti(OBu)4 ratio. The insets are the 3D reconstruction using Rhinoceront 3D software of the particles enclosed in the red square [16]. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) Please cite this article in press as: Masolo E, et al., Evaluation of the anatase/rutile phase composition influence on the photocatalytic performances…, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.05.180

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e9

to each single sample but on the other side differences in both the particle size and the pore morphology can be observed (Fig. 2B, C and D). Though the samples are highly porous, no long-range ordered pore arrangement can be perceived, which can be ascribed to a quick hydrolysis-condensation process during the gelation step and to a subsequent partial collapse of the pores during the calcination [29,34]. The powders with the highest rutile content (Fig. 2B) feature a worm-like structure with a particle size of 10e30 nm, whereas the anatase-rich sample (Fig. 2D) is characterized by smaller and defined particles with reduced degree of interconnection. In the TiO2 sample prepared with similar volumes of TiCl4/Ti(OBu)4 (1.0/0.78 ml), the morphology of the particles turns out to be different with respect to the previous ones (Fig. 2C). The same rhombohedral shape associated to a truncated bipyramide can be observed in many particles, these are however different from those of P25 in terms of size and for the presence of well-defined holes at the center of the particles (inset in Fig. 2C). These holes are attributable to the use of Pluronic 123 in the synthesis [35]. The N2 adsorptionedesorption isotherms of the titania samples calcined at 400  C and the corresponding pore size distributions are depicted in Fig. 3. The P25 material (Fig. 3A) shows a type II isotherm characteristic of macroporous materials [36,37]. A H3-type hysteresis loop is observed at high relative pressure and can be

5

related to the typical capillary condensation and evaporation processes that take place in the presence of large pores. The sample is characterized by a BET surface area of 39 ± 2 m2/g (see Table 1) and a wide pore size distribution (inset of Fig. 3A) with a maximum pore width, rBJH, of 97.5 ± 0.5 nm and cumulative pore volume of 0.47 ± 0.02 cm3/g can be observed. All the samples prepared by the acid-base pair precursors (Fig. 3B, C, D), exhibit a type IV isotherm and a H2-type hysteresis loop, characteristic of mesoporous materials (pore size between 2 and 50 nm) with ink-bottle shaped pores [38,39]. Considering that no mesostructure with long-range order is observed from the TEM images, it is possible to assume that the mesoporosity, verified by N2 physisorption measurements, is likely due to the interparticle arrangement; as well-known, there are two different types of pores, one are the interparticles pores formed from the packing of neighbouring particles and the other are the assembled pores which correspond to the position previously occupied by the surfactant P123 [40]. Pore size distributions calculated from the desorption branch of the isotherms are shown in the insets in Fig. 3B, C, D. It can be noted that, with respect to P25, the profiles of the pore size distribution are located in the mesopore region in agreement with the as discussed data. Interestingly, a narrow distribution profile is achieved for the TiO2 sample with the A:R content similar to P25 (inset in Fig. 3C): the BJH analysis presents two main peaks placed at 7.1 (the most intense) and 9.3 nm.

Fig. 3 e N2 adsorption (filled dots)-desorption (empty dots) isotherms of the commercial P25 and TiO2 samples prepared from different amount (ml) of TiCl4 and Ti(OBu)4: A. P25. B. 2.0/0.78 TiCl4/Ti(OBu)4 ratio. C. 1.0/0.78 TiCl4/Ti(OBu)4 ratio. D. 0.5/ 0.78 TiCl4/Ti(OBu)4 ratio. The insets are the BJH pore size distribution plots. Please cite this article in press as: Masolo E, et al., Evaluation of the anatase/rutile phase composition influence on the photocatalytic performances…, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.05.180

6

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e9

The pores with an average diameter size of 7.1 nm are those that are clearly distinguishable from the corresponding TEM analysis (Fig. 2C-insert 3D particle reconstruction). This aspect implies that this particular sample is characterized not only by inter-particle porosity but probably also by intra-particle porosity, not appreciable in TEM images. The estimated cumulative pore volume is 0.63 ± 0.02 cm3/g and its BET surface area corresponds to 196 ± 2 m2/g, about 5 times larger than annealed P25. It can be seen that the specific surface area of the two samples composed by the A:R weight % of 74:26 and 92:8, are 73 and 92 ± 2 m2/g, respectively, even higher than heat-treated P25, but approximately half of the m-TiO2 87:13 (A:R weight %). Hence, the large BET surface area shown by the m-TiO2 87:13 (A:R weight %) sample can be mainly ascribed to the presence of intraparticle pores obtained after the removal of P123. Fig. 4A shows a series of UV-VIS spectra collected in the wavelength (l) range between 250 and 800 nm, while Fig. 4B displays the J-V curves of the samples collected under solar irradiation in a 1M NaOH aqueous solution. The spectrum referred to the sample 1.0/0.78 ml TiCl4/Ti(OBu)4 ratio (blue curve in Fig. 4A) displays a maximum absorption at 260 nm and a shoulder at around 316 nm (lonset ¼ 410 nm), associated to the presence of both the anatase and rutile phases in the powder mixture, as also confirmed by the structural characterization [41]. The wavelength onsets and the estimated bandgap values are summarized in Table 2. For the aforementioned sample, the band gap (Eg) is calculated to be 3.02 eV, which is equal to the theoretical band gap of rutile (3.02 eV) [42]. However, since the band gap is strongly influenced by the particles size (d < 10 nm), its assignation to anatase or rutile phases may be questionable [43]. For the m-TiO2 samples characterized by the A:R weight % of 92:8 and 74:26, band gaps of 3.18 eV and 3.07 eV are calculated, which can be compared with the literature data for pure anatase (3.23 eV), and pure rutile (3.02 eV) [42e44]. The results indicate that the relative content of anatase and rutile, and hence the synthesizing parameters, play a very important role in the position of the band gap of the m-TiO2 samples. The TiO2 samples are then tested as photoanodes in the water splitting reaction (for more details see the experimental section). The photocurrent measurements are performed in 1M NaOH aqueous solution. The potential is scanned from the open circuit potential up to 1.2 V at a scan rate of 5 mV/s while irradiating the nanocrystalline films, with a radiation power of 100 W/cm2. A representative measure of the m-TiO2 is reported in Fig. 4B: the anodic current recorded when the film is exposed to the solar radiation (blue curve) results higher with respect to the current obtained during the dark test (magenta curve) for each potential applied. Additionally, the current acquired during the chopped test (red curve) alternatively matches with the singular light/dark signals. The difference between the recorded value in the presence of light and that in absence of light is used to determine the potential value at which the electrolysis has to be performed in order to verify the efficiency of the m-TiO2 photoelectrodes as photoanodes in the production of hydrogen. Fig. 4C shows the J-V curves of the films subjected to the solar irradiation. All curves are characterized by a small increase of the photocurrent density for negative potential (~1.0 V), whereas a significant increase

Fig. 4 e A. UVevisible absorbance spectra of the mesoporous TiO2 samples and P25 (HT 400  C). B. JeV curves of the m-TiO2 recorded under solar radiation (blue line), intermittent radiation (red squares), and dark test (magenta line). C. JeV curves of the mesoporous TiO2 samples and P25 (HT 400  C) under solar irradiation simulation. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

of J is observed, as expected, above þ 0.9 V. With the aim of comparing the hydrogen evolution rate of the different titania samples, electrolysis experiments are then performed at þ0.90 V and the relative data are reported in Table 2. The H2 evolution data, summarized in Table 2, show that the commercial P25 is characterized by an elevated photocatalytic activity (~10 mmol/cm2 h), confirming the literature data [45].

Please cite this article in press as: Masolo E, et al., Evaluation of the anatase/rutile phase composition influence on the photocatalytic performances…, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.05.180

7

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e9

Table 2 e Wavelength onset, bandgap values, photocurrent density data and H2 evolution rate of the as-prepared m-TiO2. Commercial P25 (HT 400  C) data are also reported for comparison. The bandgap values are calculated by the tangent method. TiCl4/Ti(OBu)4 (ml) P25 (commercial) 2.0/0.78 1.0/0.78 0.5/0.78

% Anatase

% Rutile

˚ ) anatase/rutile Crystallite size (A

Surface area BET (m2/g)

83 74 87 92

17 26 13 8

350/525 225/230 220/210 210/190

39 73ss 196 92

For the as-prepared m-TiO2 samples, the film with 87:13 anatase:rutile weight ratio resulted the most active. This sample showed a H2 evolution rate value of 4.87 mmol/cm2 h, being almost 30% more efficient than the other two systems. Considering that all the as-synthesized samples present similar crystallite sizes (see Table 1), the loss in photocatalytic activity of the other two samples (Fig. 2B and D) cannot be attributed to the anatase (more photoactive phase) content and may then be correlated with their poorer surface area. No significant difference in terms of H2 evolution rate is instead reported for these two samples. The m-TiO2 with 87:13 anatase:rutile weight ratio has a much higher surface area (196 m2/g) and a more uniform pore size, which can be associated to both a higher degree of interaction with water and also to its easier mass diffusion inside titania channels. However, this parameter alone is not sufficient to explain why the photoactivity performance of the mesoporous samples is lower when compared to the commercial P25 with a SBET of 39 m2/g. Small crystallite size accompanied by high density of microstructure defects, are considered to increase the mutual e/hþ recombination rate at the surface. In fact, as determined from the Rietveld analysis, the crystallite size of both the anatase and rutile phases in P25 calcined at 400  C are much higher (350/525 A/R) with respect to those calculated for the mesoporous samples. This implies that the crystallinity, besides the A:R ratio and the specific surface area, also plays a crucial role in the photocatalytic H2 activity. Therefore, it emerges that although the m-TiO2 with an A:R ratio of 87:13 (similar to commercial P25) presents a very interesting surface area and improved performances in respect to the other mesoporous samples, higher crystallinity (in particular for the rutile phase) is necessary to drastically decrease the number of lattice defects and then facilitate the interaction between the electrons and the water adsorbed at the TiO2 surface. This work corroborates the key role played by the crystallinity and specific surface area on the photocatalytic activity of the m-TiO2 samples, but further investigations and efforts are necessary to definitely clarify the relationship between the ordered degree of the mesoporous TiO2 and water splitting reaction performances.

Conclusions The present work provides a detailed study on the effect of the anatase-rutile microstructures in the photoelectrocatalytic production of hydrogen in mesoporous TiO2 (m-TiO2) powders. Mesoporous TiO2 (m-TiO2) specimens consisting of nanocrystalline mixed phases, i.e. with different relative

abundance of anatase ad rutile phases, have been synthesized by a facile solegel approach varying the volume ratio of the acid-base TiCl4/Ti(OBu)4 pair. The metal precursor volume ratio influences not only the final anatase and rutile relative amounts but also the specific surface area and the pore structure of the powders. On the other hand, the crystallite size of the phases only depends on the calcination temperature. All the m-TiO2 powders show higher surface area than commercial P25. Among them, the sample with an anatase:rutile weight content of 87:13, similar to commercial P25, presents uniform pore size and SBET as large as 196 m2/g, which is mainly due to the intraparticle porosity. Indeed, this sample is characterized by a promising band gap of 3.02 eV and a higher photocatalytic activity for the production of hydrogen of 4.87 mmol/cm2 h, with respect to the samples richer in either anatase or rutile phases. However, surface area is not the only parameter responsible for the enhancement of the H2 rate evolution. As demonstrated, the crystallinity degree of TiO2 also plays a crucial role in this respect.

Acknowledgements This work was funded by the University of Sassari and Regione Autonoma Sardegna by the project titled “Distribuzione di energia fuori rete (off-grids power stations): studio di materiali innovativi per il confinamento di idrogeno in stato solido” financed by “P.O.R. SARDEGNA F.S.E. 2007e2013 e  regionale e occupazione, Asse IV Obiettivo competitivita  l.3.1”. The authors gratefully Capitale umano, Linea di Attivita acknowledges Mr. Marras Antonello for the technical support  pia at UAB in the project. We thank the Servei d’ Espectrosco for their technical assistance. This work has been supported by the 2009-SGR-1292 and MAT2010-20616-C02-02 of the Generalitat de Catalunya and the Spanish MINECO, respectively. M.D.B. acknowledges financial support from an ICREAAcademia Award. E.P. acknowledges the Spanish Ministerio de Economı´a y Competitividad (MINECO) for the 'Ramon y Cajal' contract (RYC-2012-10839). The authors also thank Fondazione Banco di Sardegna for the financial support.

references

[1] Ni M, Leung MKH, Leung DYC, Sumathy K. A review and recent developments in photocatalytic water-splitting using TiO2 for hydrogen production. Renew Sustain Energy Rev 2007;11:401e25.

Please cite this article in press as: Masolo E, et al., Evaluation of the anatase/rutile phase composition influence on the photocatalytic performances…, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.05.180

8

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e9

[2] Fujishima A, Honda K. Electrochemical photolysis of water at a semiconductor electrode. Nature 1972;238:37e8. [3] Chen X, Mao SS. Titanium dioxide nanomaterials: synthesis, properties, modifications and applications. Chem Rev 2007;107:2891e959. [4] Zhang Z, Hossain MDF, Takahashi T. Photoelectrochemical water splitting on highly smooth and ordered TiO2 nanotube arrays for hydrogen generation. Int J Hydrogen Energy 2010;35:8528e35. [5] Shaislamov U, Yang BL. Single crystalline TiO2 nanorods with enhanced visible light activity for solar hydrogen generation. Int J Hydrogen Energy 2013;38:14180e8. [6] Fujishima A, Rao TN, Tryk DA. Titanium dioxide photocatalysis. J Photochem Photobiol C Photochem Rev 2000;1:1e21. [7] Chen C, Ma W, Zhao J. Semiconductor-mediated photodegradation of pollutants under visible-light irradiation. Chem Soc Rev 2010;39:4206e19. [8] Fox MA, Dulay MT. Heterogeneous photocatalysis. Chem Rev 1993;93:341e57. [9] Chen HW, Liao YT, Wu KCW, Ho KC. Fabrication and characterization of plastic-based flexible dye-sensitized solar cells consisting of crystalline mesoporous titania nanoparticles as photoanodes. J Mater Chem 2011;21:17511e8. [10] Luo H, Wang C, Yan Y. Synthesis of mesostructured titania with controlled crystalline framework. Chem Mater 2003;15:3841e6. [11] Tong T, Zhang J, Tian B, Chen F, He D. Preparation and characterization of anatase TiO2 microspheres with porous framework via controlled hydrolysis of titanium alkoxide followed by hydrothermal treatment. Mater Lett 2008;62:2970e2. [12] Zhang YY, Chen JZ, Li XJ. Preparation and photocatalytic performance of anatase/rutile mixed-phase TiO2 nanotubes. Catal Lett 2010;139:129e33. [13] Ohno T, Sarukawa K, Tokieda K, Matsumura M. Morphology of a TiO2 photocatalyst (Degussa, P-25) consisting of anatase and rutile crystalline phases. J Catal 2011;203:82e6. [14] Zhang J, Xu Q, Feng Z, Li M, Li C. Importance of the relationship between surface phases and photocatalytic activity of TiO2. Angew Chem Int Ed 2008;47:1766e9. [15] Hong Z, Dai H, Huang Z, Wei M. Shape-controlled TiO2 nanoparticles and TiO2 P25 interacting with CO and H2O2 molecular probes: a synergic approach for surface structure recognition and physico-chemical understanding. Phys Chem Chem Phys 2013;35:307e15. [16] Hall RN. Electron-hole recombination in germanium. Phys Rev 1952;87:387. [17] Zhang Z, Yates JT. Direct observation of surface-mediated electronhole pair recombination in TiO2(110). J Phys Chem C 2010;114:3098e101. [18] Li H, Shi J, Liang J, Li X, Li L, Ruan M. Synthesis of well ordered mesoporous titania powder with crystallized framework. Mater Lett 2008;62:1410e3. [19] Tian B, Liu X, Tu B, Yu C, Fan J, Wang L, et al. Self-adjusted synthesis of ordered stable mesoporous minerals by acidbase pairs. Nat Mater 2003;2:159e63. [20] O'Reagan B, Gratzel M. A low cost, high-efficiency solar cell based on dye sensitized colloidal TiO2 films. Nature 1991;353:737e40. [21] Vivero-Escoto JL, Chiang YD, Wu KCW, Yamauchi Y. Recent progress in mesoporous titania materials: adjusting morphology for innovative applications. Sci Technol Adv Mater 2012;13:013003e11. [22] Chen L, Yao B, Cao Y, Fan K. Synthesis of well-ordered mesoporous titania with tunable phase content and high photoactivity. J Phys Chem C 2007;111:11849e53.

[23] Wan Y, Zhao DY. On the controllable soft-templating approach to mesoporous silicates. Chem Rev 2007;13:2821e60. [24] Zhang J, Deng Y, Gu D, Wang S, She L, Che R, et al. Ligandassisted assembly approach to synthesize large-pore ordered mesoporous titania with thermally stable and crystalline framework. Adv Energy Mater 201; 1: 241e248. [25] Mahoney L, Koodali RT. Versatility of evaporation-induced self-assembly (EISA) method for the preparation of mesoporous TiO2 for energy and environmental applications. Materials 2014;7:2697e746. [26] Brinker CJ, Lu Y, Sellinger A, Fan H. Evaporation-induced self-assembly: nanostructures made easy. Adv Mater 1999;11:579e85. [27] Choi SY, Mamak M, Coombs N, Chopre N, Ozin GA. Thermally stable two-dimensional hexagonal mesoporous nanocrystalline anatase, Meso-nc-TiO2: bulk and crackfree thin film morphologies. Adv Funct Matters 2004;14:335e44. [28] Wang CC, Ying J. Sol-Gel synthesis and hydrotermal processing of anatase and rutile titania nanocrystals. Chem Mater 1999;11:3113e20. [29] Crepaldi EL, Soler-Illa GJ de AA, Grosso D, Cagnol F, Ribot F, Sanchez C. Controlled formation of highly organized titania thin films: from mesostructure hybrids to mesoporous nanoanatase TiO2. J Am Chem Soc 2003;125:9770e86. [30] Deiana C, Minella M, Tabacchi G, Maurino V, Fois E, Martra G. Shape-controlled TiO2 nanoparticles and TiO2 P25 interacting with CO and H2O2 molecular probes: a synergic approach for surface structure recognition and physicochemical understanding. Phys Chem Chem Phys 2013;15:307e15.  MD, [31] Masolo E, Meloni M, Garroni S, Mulas G, Enzo S, Baro et al. Mesoporous titania powders: the role of precursors, ligand addition and calcination rate on their morphology, crystalline structure and photocatalytic activity. Nanomaterials 2014;4:583e98. [32] Scardi P, Lutterotti L, Maistrelli P. Experimental determination of the instrumental broadening in the BraggBrentano geometry. Powder Diffr 1994;9:180e6. [33] Bard AJ, Faulkner LR. Electrochemical methods e fundamentals and applications. 2nd ed. New York: Wiley; 2001. [34] Schu¨t F. Non siliceous mesostructured and mesoporous materials. Chem Mater 2001;13:3184e95. [35] Wan Y, Yang H, Zhao D. “Host-guest” chemistry in the synthesis of ordered nonsiliceous mesoporous materials. Accounts Chem Res 2006;37:423e32. [36] Sing KSW, Everett DH, Haul RAW, Moscou L, Pierotti RA, Rouquerol J, et al. Reporting physisorption data for gas/solid with special reference to the determination of surface area and porosity. Pure Appl Chem 1985;57:603e19. [37] Thomas JM, Thomas WJ. Principles and practice of heterogeneous catalysis. Weinheim: VCH; 1997. [38] Zhao D, Wan Y, Zhou W. Ordered mesoporous materials. Weinheim, Germany: Wiley-VCH; 2013. [39] Parmar KPS, Ramasamy E, Leeb JW, Sung Lee J. A simple method for producing mesoporous anatase TiO2 nanocrystals with elevated photovoltaic performance. Scr Mater 2010;62:223e6. [40] Jing W, Huang W, Xing W, Wang Y, Jin W, Fan Y. Fabrication of supported mesoporous TiO2 membranes: Matching the assembled and interparticle pores for an improved ultrafiltration performance. ACS Appl Mater Interfaces 2009;1:1607e12. [41] Kho YK, Iwase A, Teoh WT, Madler L, Kudo A, Amal R. Photocatalytic H2 evolution over TiO2 nanoparticles. The

Please cite this article in press as: Masolo E, et al., Evaluation of the anatase/rutile phase composition influence on the photocatalytic performances…, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.05.180

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 5 ) 1 e9

synergistic effect of anatase and rutile. J Phys Chem C 2010;114:2821e9. [42] Rao MV, Rajeshwar K, Pal Verneker VR, DuBow JJ. Photosynthetic production of hydrogen and hydrogen peroxide on semiconducting oxide grains in aqueous solutions. J Phys Chem 1980;84:1987. [43] Wang Y, Herron N. Nanometers-sized semiconductor clusters: materials synthesis, quantum size effects, and photophysical properties. J Phys Chem 1991;95:525e32.

9

[44] Hoffman MR, Martin ST, Choi W, Bahnemann DW. Environmental applications of semiconductor photocatalysis. Chem Rev 1995;95:525e32. [45] Bojinova A, Kralchevska R, Poulios I, Dushkin C. Anatase/ rutile TiO2 composites: influence of the mixing ratio on the photocatalytic degradation of malachite green and orange II in slurry. Mater Chem Phys 2007;106:187.

Please cite this article in press as: Masolo E, et al., Evaluation of the anatase/rutile phase composition influence on the photocatalytic performances…, International Journal of Hydrogen Energy (2015), http://dx.doi.org/10.1016/j.ijhydene.2015.05.180