- Email: [email protected]
Enhancing the photoelectrochemical water splitting activity of rutile nanorods by removal of surface hydroxyl groups
Catalysis Today 259 (2016) 360–367
Contents lists available at ScienceDirect
Catalysis Today journal homepage: www.elsevier.com/locate/cattod
Enhancing the photoelectrochemical water splitting activity of rutile nanorods by removal of surface hydroxyl groups Shaohong Jiang, Yang Li, Xiaoli Zhang, Yongdan Li ∗ Collaborative Innovation Center of Chemical Science and Engineering (Tianjin); Tianjin Key Laboratory of Applied Catalysis Science and Technology; State Key Laboratory of Chemical Engineering (Tianjin University); School of Chemical Engineering, Tianjin University, Tianjin 300072, China
a r t i c l e
i n f o
Article history: Received 2 February 2015 Received in revised form 13 May 2015 Accepted 24 May 2015 Keywords: Rutile nanorods Surface Hydroxyl groups Photoelectrochemical water splitting Hydrogen
a b s t r a c t The effect of the surface hydroxyl (OH) groups on the photoelectrochemical (PEC) water splitting performance of rutile nanorod is examined. Based on the results of characterization, the surface OH groups are demonstrated to serve as the recombination centers for the photo-generated charges and to hinder the charge transfer at the interfaces. H2 O2 thermal and anneal treatments are applied to reduce the OH group surface coverage and both are proved effective. With reducing the relative surface OH coverage from 47 to 8%, the photocurrent density is increased from 0.419 to 0.642 mA cm−2 at 0 V (vs. Ag/AgCl), achieving 53.2% improvement. Compared with the thermal treatment, H2 O2 treatment is more effective and energy-efficient. This work highlights that the surface property of the material itself is an important factor in the design of the PEC electrodes for water splitting. © 2015 Elsevier B.V. All rights reserved.
1. Introduction TiO2 photoanode has been intensively investigated for photoelectrochemical (PEC) water splitting because of its high chemical stability, nontoxicity, and low cost [1,2]. PEC water splitting includes charge transfer steps across the semiconductor/semiconductor and semiconductor/aqueous interfaces, and surface reaction steps, which involves charge transfer and reaction. Therefore, the efficiencies of photo-generated charge separation and transfer between the interfaces are the key factors deciding the overall efficiency of the process [3–5]. Therefore, the surface modification techniques, such as co-catalysts loading and semiconductors combination, have been intensively investigated for TiO2 to enhance the photo-generated charge separation and transfer in water splitting reaction. With respect to the surface property of TiO2 itself, it has been known that the surface hydroxyl (OH) groups play an important role in the adsorption of O2 and affect the overall water splitting activity [6]. Oosawa and Grätzel and Kobayakawa et al. [7–9] investigated the effect of the surface OH density on the photocatalytic O2 evolution reaction over a rutile TiO2 powder catalyst from AgNO3 solution. Oosawa and Grätzel [7,8] found that the O2 generation activity is inversely proportional to the density of the surface OH groups. Kobayakawa et al. [9] reported that the O2
∗ Corresponding author. Tel.: +86 22 27405613; fax: +86 22 27405243. E-mail address: [email protected] (Y. Li). http://dx.doi.org/10.1016/j.cattod.2015.05.022 0920-5861/© 2015 Elsevier B.V. All rights reserved.
generation activity achieved a maximum under an optimal surface OH density. In these works, the authors ascribed the effect of the OH groups to the facilitated charge recombination on the surface OH sites. Though more direct and quantitative evidence is needed, these works show the importance of the material surface property in the design of the photocatalysts for water splitting. Recently, many works focused on one-dimensional nanostructures, because of the improvement in charge carriers separation and transport on such materials [10–13]. In 2009, Liu and Aydil [14] reported a hydrothermal growth preparation method for the oriented single-crystalline rutile TiO2 nanorods on fluorine-doped tin oxide (FTO) substrate, since then many works to enhance the PEC water splitting activity utilizing the rutile nanorods have been published, e.g. CdS and Cu2 S quantum dots-sensitization, CdS/CdSe and ln2 S3 /AgInS2 co-sensitization, as well as the branched and threedimensional hierarchical nanostructures utilizations [15–20]. It is noted that in the previous reports on the effects of the surface OH groups, rutile powder was dispersed in the solution and was utilized for the O2 evolution reaction. The understanding on the effect of surface OH groups on the PEC water splitting activity of the nanorods is meaningful. With respect to the removal of the surface OH groups, anneal treatment was practiced, which requires high temperature and is accompanied with the risks of phase transformation and sintering. H2 O2 treatment is recently used for the catalytic material surface modification. In this work, rutile TiO2 nanorods are used as the photoanode material for PEC water splitting. The role of the surface OH groups
S. Jiang et al. / Catalysis Today 259 (2016) 360–367
361
on the catalytic performance in terms of the separation and transfer of the photo-generated electrons and holes is elucidated. H2 O2 treatment (HTT) is utilized as a tool to monitor the coverage of the surface OH group. As comparison, a regular anneal treatment (AT) is also discussed. 2. Material and methods 2.1. Preparation of rutile nanorod photoanode Rutile nanorods were directly grown on a FTO substrate with a hydrothermal process [14]. Hydrochloric acid (37% by weight), titanium butoxide (AR, Tianjin Guangfu Technology Development Co., Ltd) were used as the raw materials. In a typical synthesis, two pieces of FTO substrates (3×3 cm2 ), which had been ultrasonically cleaned for 30 min in a mixed solution of deionized water, acetone and 2-propanol with volume ratios of 1:1:1, were placed at an angle against the wall of a teflon lined (100 mL volume) stainless steel autoclave with the conductive side facing down. Deionized water (30 mL) and hydrochloric acid (30 mL) was mixed and ultrasonically stirred for 5 min, and then 1 mL of titanium butoxide was added. After another 5 min of ultrasonic stirring, the solution was transferred into the autoclave. Then the autoclave was sealed and maintained at 180 ◦ C for 4 h followed by natural cooling to room temperature. The as-grown samples were rinsed with deionized water and absolute ethanol, dried in ambient air, and then annealed at 450 ◦ C for 0.5 h to enhance the contact between the rutile nanorods and FTO layer. The electrode obtained after above steps is referred to as the raw sample. 2.2. Modification of surface coverage of hydroxyl groups AT and HTT were carried out, respectively, to monitor the surface coverage of the OH groups. Samples after AT at 450 ◦ C for 1, 2, 3 and 4 h are referred to as AT-xh (AT, anneal treatment; x, treatment time). In HTT, a piece of raw sample was put into a 100 mL beaker at an angle against the wall with the conductive side facing down. After added 50 mL H2 O2 solution (30% by weight, Tianjin Guangfu Technology Development Co., Ltd) without further treatment, the beaker was placed into the water bath with the temperature maintained at 80 ◦ C for 0.5, 1, 1.5 and 2 h. Samples after HTT are referred to as HTT-xh (HTT, H2 O2 thermal treatment; x, treatment time). In addition, H2 O2 treatment was also carried out at 25 ◦ C for 0.5 and 1 h, and the samples after this are referred to as HNT-xh. 2.3. Characterization The powder X-ray diffraction (XRD) patterns were recorded with a Bruker D8 Advance X-ray Diffractometer with Cu K␣ radiation. The surface morphology was observed with a Hitachi S-4800 field-emission scanning electron microscope (SEM). The high-resolution transmission electron microscopy (HR-TEM) analysis was carried out with a JEOL JEM-2100F instrument. The UV−vis absorption spectra of the samples were acquired with an UV−vis spectrometer (PerkinElmer, Lambda 750S). The Photoluminescence (PL) spectra were obtained with Jobin Yvon Fluorolog 3-21 in 1 M NaOH solution to identify the photo-generated charges separation and utilization efficiency. The surface properties of the prepared samples were characterized by X-ray photoelectron spectroscopy (XPS, Perkinelmer, PHI1600 ESCA). 2.4. Photoelectrochemical water splitting measurements The photocurrent density (J)−potential (V) and chopped J−time (t) curves were obtained with a three-electrode setup, in which we used rutile nanorods as the working, Pt gauze as the counter,
Fig. 1. (a) X-ray diffraction patterns of the bare FTO substrate and rutile nanorods arrays with/without surface modifications, (b) UV−vis light absorption spectra of the raw sample, HTT-xh and AT-xh samples.
Ag/AgCl as the reference electrodes, and 1 M NaOH solution as the electrolyte. A 300 W Xe-lamp (Perfect Light, PLS-SXE-300) coupled with an AM 1.5G filter was applied as the light source. By changing the applied lamp current and the distance between the lamp and the rutile nanorods photoanode, the illumination intensity was adjusted to 100 mW cm−2 . The PEC data were measured using a Princeton Applied Research Versa STAT3 Potentiostat. The J−V curves were obtained with a scan rate of 5 mV s−1 , and the chopped J−t curves were obtained at an applied potential of 0 V (vs. Ag/AgCl). Electrochemical impedance spectroscopy (EIS) was measured with applying a 10 mV AC signal within a frequency range from 0.5 Hz to 100 kHz at the open-circuit potential of the PEC cells. 3. Results and discussion Fig. 1a displays the XRD patterns of the bare FTO substrate and rutile nanorods grown on FTO. The diffraction peaks of the grown samples agree well with the tetragonal rutile phase (PDF no. 88−1175), and the (1 0 1) and (0 0 2) planes can be assigned. The (0 0 2) diffraction peak appears as the most intense one,
362
S. Jiang et al. / Catalysis Today 259 (2016) 360–367
Fig. 2. SEM of the raw sample (a), HTT-2 h (b), AT-4 h (c) and HR-TEM of the raw sample.
indicating the oriented growth of rutile film. Comparing the patterns of the samples with HTT and AT at different conditions, no significant influence of HTT and AT on the crystal structure was observed. The UV−vis light absorption spectra of all the samples are shown in Fig. 1b. It is obvious that the spectra of the raw rutile nanorods film, HTT-xh and AT-xh samples are almost identical, and consistently show the onset wavelength at 413 nm (3.0 eV) agreeing with that reported in literature [1,21]. It indicates that HTT and AT have little effect on the UV−vis light absorption ability of the rutile nanorod photoanode. Fig. 2a−c shows that the rutile nanorods are tetragonal in shape. Comparing the morphology of the raw sample to that of HTT-2 h and AT-4 h that took the longest treatment time, no obvious change is noticed, which indicates that HTT and AT have no influence on the morphology of rutile nanorods. Lattice fringes of (0 0 1) and (1 1 0) planes with interplanar spacings of 0.298 and 0.322 nm, respectively, can be assigned in Fig. 2b. Plane of (0 0 1) is perpendicular to the growth direction, while that of (1 1 0) is parallel to it. This suggests that the rutile nanorods grew in the [001] direction, which is consistent with the XRD analysis and the previous results of Liu [14]. To obtain the detailed information of surface chemical states, X-ray photoelectron spectroscopy (XPS) measurements were carried out. In term of Ti 2p XPS spectra (Fig. 3a), the two main peaks, located at 458.6 and 464.4 eV, are assigned to Ti4+ 2p3/2 and Ti4+ 2p1/2 of TiO2 [22]. It is reported that the shoulder peaks at 457.5
and 463.2 eV correspond to Ti3+ 2p3/2 and Ti3+ 2p1/2 are attributed to the Ti3+ state that serves as the photo-generated charges recombination center [22,23]. To figure out the effect of HTT and AT on the Ti3+ state concentration, the relative atomic concentrations of Ti4+ 2p3/2 and Ti3+ 2p3/2 of all the samples examined are calculated, and remain identical as summarized in Table 1. This indicates that HTT and AT did not affect the Ti3+ state concentration. It is also noted here that the impurity of Cl was not detected by XPS, and that HTT and AT have little influence on the removal of the organic impurities (Supporting Information 1, SI 1).
Table 1 Surface chemical composition of all the samples measured by XPS (relative atomic concentration, %). Samples
OL
OH
Ti4+ 2p3/2
Ti3+ 2p3/2
Raw HTT-0.5 h HTT-1 h HTT-1.5 h HTT-2 h AT-1 h AT-2 h AT-3 h AT-4 h HNT-0.5 h HNT-1 h
53 65 85 92 92 59 76 84 85 56 55
47 35 15 8 8 41 24 16 15 44 45
92 92 92 92 92 92 92 92 92 92 92
8 8 8 8 8 8 8 8 8 8 8
S. Jiang et al. / Catalysis Today 259 (2016) 360–367
363
Fig. 3. X-ray photoelectron spectra of Ti 2p (a) and O 1s (b) core levels over the raw sample, HTT-xh and AT-xh.
In term of the O 1s XPS spectra (Fig. 3b), two peaks referring to two kinds of oxygen species are observed. The intense peak at about 529.97 eV is attributed to the oxygen in the TiO2 crystal lattice (OL ), while that at 531.98 eV is assigned to the surface OH [24–27]. It is obvious that the intensity of OL showed little dependence on the HTT and AT conditions, but that of the surface OH reduced along with the increase of the treatment time for each surface modification technique. This suggests that HTT and AT have little influence on the crystal structure of the rutile nanorods, which is also supported by the data of XRD, but they change the surface coverage of the OH groups. As presented in Table 1, the relative surface percentage of the OH groups of the raw sample, HTT-0.5 h, HTT-1 h, HTT-1.5 h, HTT-2 h, AT-1 h, AT-2 h, AT-3 h and AT-4 h are 47, 35, 15, 8, 8, 41, 24, 16 and 15%, respectively. With respect to each surface modification technique, the relative surface OH group coverage reduces along with the increase of the treatment time, and achieves the minimum level at 1.5 and 3 h for HTT and AT, respectively. The AT at the same temperature of HTT (80 ◦ C) was also carried out (SI 2). The result demonstrates that this temperature is not high enough to remove the surface OH groups. Therefore, under conditions of less energy intensive and shorter treatment time, HTT gives a better OH groups removal efficiency.
The PEC water splitting activity of the rutile nanorods arrays with or without HTT modification was measured with a threeelectrode electrochemical configuration in 1 M NaOH electrolyte under AM 1.5G (100 mW cm−2 ) illumination. Fig. 4a shows the J−V curves obtained by linear sweep voltammetry measurements. The photocurrent densities of the raw sample, HTT-0.5 h, HTT-1 h, HTT-1.5 h, HTT-2 h, AT-1 h, AT-2 h, AT-3 h and AT-4 h are 0.419, 0.514, 0.588, 0.642, 0.642, 0.491, 0.556, 0.587 and 0.587 mA cm−2 , respectively, at 0 V (vs. Ag/AgCl). Compared with the raw sample, HTT-0.5 h, HTT-1 h, HTT-1.5 h and HTT-2 h achieve 22.7, 40.3, 53.2 and 53.2% photocurrent density enhancement, while AT1 h, AT-2 h, AT-3 h and AT-4 h achieve 17.2, 32.7, 40.1 and 40.1% enhancement, respectively. Considering the surface OH in general, the photocurrent density increases with respect to the reduction of the relative surface OH group coverage. It indicates that the removal of surface OH groups over rutile nanorods is beneficial for PEC water splitting. The photo response and PEC water splitting stability are measured with the chopped J−t curves at 0 V (vs. Ag/AgCl) as plotted in Fig. 4b. The transient change of photocurrent density at each light-on and light-off moment demonstrates the fast photo response of the samples. A constant photocurrent density of each sample is obtained without decay in the light-on
364
S. Jiang et al. / Catalysis Today 259 (2016) 360–367
Fig. 4. PEC water splitting activity measured with a three-electrode setup in NaOH solution (1 M). (a,b) J−V curves with a scan rate of 5 mV s−1 from -0.8 to 0.5 V (vs. Ag/AgCl). (c) Chopped J−t curves at an applied potential of 0 V (vs. Ag/AgCl).
region of every cycle and shows stability through 5000s testing. This demonstrates that the J−t curves are highly reproducible and the PEC water splitting activity is stable for all the samples. PL spectrum is a facial and effective way to measure the separation and the utilization of the photo-generated electrons and holes [28,29]. The more intense the peak shows in the PL spectrum, the less separation and utilization efficiency the photo-generated charges possess. As plotted in Fig. 5, an intense peak at 427 nm appearing for all the samples is attributed to the recombination between holes trapped by surface OH groups and electrons [30]. The monotonous decrease of the intensity of this peak with respect to the reduction of the relative surface OH group coverage and the increase of the photocurrent density can be observed. As mentioned above, neither the structure, UV−vis light absorption nor Ti3+ state concentration of the rutile nanarods were affected by the HTT and AT, and no impurity was detected by XPS. The only significant difference appeared is the amount of the surface OH groups on the rutile nanorods. Therefore, all samples with PL testing have the identical bulk quality, and PL spectrum indicate the difference of charges seperation caused by the various surface OH groups coverage. Better separation and more effective utilization of the photo-generated electrons and holes can be achieved by removing the surface OH
Fig. 5. Photoluminescence spectra of the raw sample, HTT-xh and AT-xh excited at = 330 nm.
S. Jiang et al. / Catalysis Today 259 (2016) 360–367
Scheme 1. Schematic illustration of the surface OH groups on the rutile nanorods array electrode, and the recombination and transfer impediment of the photogenerated charges caused by surface OH.
groups, resulting in the enhancement of the PEC water splitting activity. The mechanism of recombination process caused by the surface OH is then verified. With respect to the photocatalytic degradation of the organic pollutant, such as methyl orange and phenol, over rutile, it is generally demonstrated and accepted that the photogenerated holes can be easily captured by the surface OH groups to produce hydroxyl radicals (Reaction 1), which possess strong oxidation ability and would trigger the subsequent degradation steps [31–34]. It is rational to assume that this reaction occurs on the rutile nanorod surface covered with OH groups. It is reported that there are three kinds of surface hydroxyl groups, Ti-(OH2 + ), Ti(OH) and Ti-(O− ), with their proportion depending on the solution pH [35–37]. In NaOH solution, surface hydroxyl is deprotonated as Ti-(O− ).
˙ Ti − O− + h+ → Ti − (O)
(1)
The hydroxyl radicals capture the photo-generated electrons, leading to the recombination (Reaction 2 and Scheme 1) and the decrease of the number of the available electrons and holes for hydrogen and oxygen evolution.
Ti − O˙ + e− → Ti − O− + hv
(2)
We agree with the suggestions of Rothenberger et al. [38] and Brown and Darwen [39] that holes are the minority carriers and should be removed rapidly from TiO2 . Co-catalysts, such as RuO2 and IrO2 , loaded on the host catalyst surface facilitates to achieve this purpose since they serve as the holes transfer catalysts. As plotted in Fig. 6, the typical EIS are presented as Nyquist plots and utilized to measure the surface charge transfer resistance. It is observed that with the reduce of the surface OH group coverage, the semicircle of the plot becomes smaller, which indicates the effective enhancement of charge transfer [40–42]. This suggests that the OH groups serve as the charge transfer obstacles over the rutile nanorod surface (Scheme 1). Taking reaction 1 into consideration, the OH groups separate the rutile nanorods and the electrolyte, and hinder the transfer of holes between these interfaces for O2 production, leading to the increase of photo-generated charges recombination probability and the ultimate reduction of PEC water splitting activity. The surface OH groups serve in two ways: (i) acting as the recombination centers for the photo-generated charges, (ii) hindering the charges transfer. These two disadvantages contribute to the degradation of the PEC performance over the rutile nanorods photoanode. In term of the recombination, the non-radiative process is the major part, and the efficiency of it is associated with
365
Fig. 6. Electrochemical impedance spectra presented as Nyquist plots of the raw sample, HTT-xh and AT-xh.
structural defects, impurities and surface trap sites [43–45]. Removal of surface OH groups may also be benefit for the inhibition of the non-radiative recombination. Further investigation on the contribution between recombination (radiative and nonradiative) and the charges transfer impediment to the degradation of PEC performance is needed in the future. To figure out the detail of the surface OH groups removal by HTT, we made a comparison experiment in which the H2 O2 treatment was done at 25 ◦ C (referred to as HNT), and compared the removal rates of the surface OH groups between HTT and HNT. The relative surface percentage of the OH groups of HNT-0.5 h and HNT-1 h are 44 and 45% (Fig. 7 and Table 1). Within the same treatment time, the surface OH group removal effect of HNT is much less efficient than that of HTT. The major difference between the two conditions is the H2 O2 decomposition rate, depending on the treatment temperature. The interaction between H2 O2 and the metal oxide surfaces was investigated in the context of nuclear technology, since H2 O2 is one of the major radiolysis products that drive corrosion of the nuclear facilities. In these precious works, it has been confirmed that the H2 O2 decomposition in the presence of metal oxide, such as TiO2 and ZrO2 , is a surface radical reaction [46,47]. The initial decomposition steps involve the molecule adsorption (Reaction 3) and the cleavage of the O O bond for hydroxyl radical production (Reaction 4), which is temperature-dependent. Therefore, more hydroxyl radicals can be produced with the increase of the treatment temperature from 25 to 80 ◦ C. In terms of the rutile nanorod surface covered with OH groups, it is proposed that the hydroxyl radicals substitute the surface OH groups forming adsorption state because of its strong bonding with the metal cation [46], resulting in the removal of the surface OH groups (Reaction 5). The subsequent H2 O2 decomposition processes are presented as the general mechanism in reactions 6−8. H2 O2 (aq) → H2 O2 (ads)
(3)
H2 O2 (ads) → 2 · OH(aq)
(4)
˙ ˙ 2 · OH (aq) + 2Ti-OH → 2 · OH (ads) /(HO) Ti O Ti (OH) + H2 O
(5)
·OH (ads) + H2 O2 (ads) → H2 O + ·OOH(ads)
(6)
·OH (ads) + ·OOH (ads) → O2 ↑ +H2 O
(7)
2 · OOH (ads) → O2 ↑ +H2 O2 (ads)
(8)
366
S. Jiang et al. / Catalysis Today 259 (2016) 360–367
Fig. 7. X-ray photoelectron spectra of (a) Ti 2p and (b) O 1s core levels over HNT-xh.
4. Conclusion With respect to the surface OH groups removal, H2 O2 treatment is more effective and energy-efficient than anneal treatment. The temperature dependence of the H2 O2 treatment indicates that the attack of hydroxyl radical produced from H2 O2 decomposition is the major reason for the removal of surface OH groups. In terms of the role of the surface OH groups on rutile nanorods for PEC water splitting, it is demonstrated that the surface OH groups serve as the recombination centers of the photo-generated electrons and holes, as well as hinder the direct transfer of the holes to the free hydroxyl ions. The PEC water splitting activity is enhanced by removing the surface OH groups. Reducing the relative surface OH coverage from 47 to 8%, the photocurrent density is enhanced by 53.2% at 0 V (vs. Ag/AgCl). To figure out the major degradation between the recombination and hindering effects over the surface OH groups, further study is needed in the future. This work highlights the crucial importance of the material surface property itself in the design of the PEC electrodes for water splitting. Acknowledgement This work has been supported in part by the Program of Introducing Talents to the University Disciplines under file number B06006, and the Program for Changjiang Scholars and Innovative Research Teams in Universities under file number IRT 0641. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.cattod.2015.05.022. References [1] [2] [3] [4] [5]
A. Fujishima, K. Honda, Nature 238 (1972) 37–38. B. O’Regan, M. Grätzel, Nature 353 (1991) 737–740. X. Chen, S. Shen, L. Guo, S.S. Mao, Chem. Rev. 110 (2010) 6503–6570. J. Yang, D. Wang, H. Han, C. Li, Acc. Chem. Res. 46 (2013) 1900–1909. X. Zhang, J. Zhao, S. Wang, H. Dai, X. Sun, Int. J. Hydrogen Energ. 39 (2014) 8238–8245.
[6] G. Munuera, V. Rives-Arnau, A. Saucedo, J. Chem. Soc. Faraday Trans. 1 Phys. Chem. Condens. Phases 75 (1979) 736–747. [7] Y. Oosawa, M. Grätzel, J. Chem. Soc. Chem. Commun. (1984) 1629–1630. [8] Y. Oosawa, M. Grätzel, J. Chem. Soc. Faraday Trans. 1 Phys. Chem. Condens. Phases 84 (1988) 197–205. [9] K. Kobayakawa, Y. Nakazawa, M. Ikeda, Y. Sato, A. Fujishima, Berichte Der Bunsengesellschaft Für Phys. Chemie. 94 (1990) 1439–1443. [10] M. Law, L.E. Greene, J.C. Johnson, R. Saykally, P. Yang, Nat. Mater. 4 (2005) 455–459. [11] T. Lindgren, H. Wang, N. Beermann, L. Vayssieres, A. Hagfeldt, S.E. Lindquist, Sol. Energ. Mater. Sol. C 71 (2002) 231–243. [12] M. Liu, N. de Leon Snapp, H. Park, Chem. Sci. 2 (2011) 80. [13] A. Wolcott, W.A. Smith, T.R. Kuykendall, Y. Zhao, J.Z. Zhang, Small 5 (2009) 104–111. [14] B. Liu, E.S. Aydil, Am. Chem. Soc. 131 (2009) 3985–3990. [15] I.S. Cho, Z. Chen, A.J. Forman, D.R. Kim, P.M. Rao, T.F. Jaramillo, Nano Lett. 11 (2011) 4978–4984. [16] K. Guo, Z. Liu, C. Zhou, J. Han, Y. Zhao, Z. Liu, Appl. Catal. B Environ. 154–155 (2014) 27–35. [17] Y.L. Lee, C.F. Chi, S.Y. Liau, Chem. Mater. 22 (2010) 922–927. [18] Z. Liu, K. Guo, J. Han, Y. Li, T. Cui, B. Wang, Small 10 (2014) 3153–3161. [19] H. Wang, Y. Bai, H. Zhang, Z. Zhang, J. Li, L. Guo, J. Phys. Chem. C 114 (2010) 16451–16455. [20] J.S. Yang, W.P. Liao, J.J. Wu, ACS Appl. Mater. Interfac. 5 (2013) 7425–7431. [21] R.I. Bickley, T. Gonzalez-Carreno, J.S. Lees, L. Palmisano, R.J.D. Tilley, J. Solid State Chem. 92 (1991) 178–190. [22] J. Zhao, J. Yao, Y. Zhang, M. Guli, L. Xiao, J. Power Sources 255 (2014) 16–23. [23] X. Zhang, H. Tian, X. Wang, G. Xue, Z. Tian, J. Zhang, Mater. Lett. 100 (2013) 51–53. [24] A. Naldoni, M. Allieta, S. Santangelo, M. Marelli, F. Fabbri, S. Cappelli, J. Am. Chem. Soc. 134 (2012) 7600–7603. [25] B. Santara, P.K. Giri, K. Imakita, M. Fujii, Nanoscale 5 (2013) 5476–5488. [26] J.C. Yu, J. Yu, H.Y. Tang, L. Zhang, J. Mater. Chem. 12 (2002) 81–85. [27] J. Zhuang, S. Weng, W. Dai, P. Liu, Q. Liu, J. Phys. Chem. C 116 (2012) 25354–25361. [28] Y. Li, X. Zhang, S. Jiang, H. Dai, X. Sun, Y. Li, Sol. Energ. Mater. Sol. C 132 (2015) 40–46. [29] X. Zhang, Y. Li, J. Zhao, S. Wang, Y. Li, H. Dai, J. Power Sources 269 (2014) 466–472. [30] S. Mathew, A.K. Prasad, T. Benoy, P.P. Rakesh, M. Hari, T.M. Libish, J. Fluoresc. 22 (2012) 1563–1569. [31] X. Fu, L.A. Clark, Q. Yang, M.A. Anderson, Environ. Sci. Technol. 30 (1996) 647–653. [32] P. Salvador, Prog. Surf. Sci. 86 (2011) 41–58. [33] A. Sclafani, L. Palmisano, J. Phys. Chem. 94 (1990) 829–832. [34] S.-J. Tsai, S. Cheng, Catal. Today 33 (1997) 227–237. [35] N. Jaffrezic-Renault, P. Pichat, A. Foissy, R. Mercier, J. Phys. Chem. 90 (1986) 2733–2738. [36] S.I. Nishimoto, B. Ohtani, H. Kajiwara, T. Kagiya, J. Chem. Soc. Faraday Trans. Phys. Chem. Condens. Phases 79 (1983) 2685–2694.
S. Jiang et al. / Catalysis Today 259 (2016) 360–367 [37] A.P. Norton, S.L. Bernasek, A.B. Bocarsly, J. Phys. Chem. 92 (1988) 6009–6016. [38] G. Rothenberger, J. Moser, M. Grätze, N. Serpone, D.K. Sharma, J. Am. Chem. Soc. 107 (1985) 8054–8059. [39] G.T. Brown, J.R. Darwent, J. Phys. Chem. 88 (1984) 4955–4959. [40] H. Zhang, X. Lv, Y. Li, Y. Wang, J. Li, ACS Nano 4 (2010) 380–386. [41] J. Meng, Z. Yu, Y. Li, Y. Li, Catal. Today 225 (2014) 136–141. [42] Z. Yu, J. Meng, Y. Li, Y. Li, Int. J. Hydrogen Energ. 38 (2013) 16649–16655.
367
[43] R. Sellappan, A. Galeckas, V. Venkatachalapathy, A.Y. Kuznetsov, D. Chakarov, Appl. Catal. B: Environ. 106 (2011) 337–342. [44] J. Buha, Thin Solid Films 545 (2013) 234–240. [45] B. Choudhury, A. Choudhury, Phys. E Low-Dimensional Syst. Nanostruct. 56 (2014) 364–371. [46] M. Jonsson, Isr. J. Chem. 54 (2014) 292–301. [47] C.M. Lousada, A.J. Johansson, T. Brinck, M. Jonsson, J. Phys. Chem. C 116 (2012) 9533–9543.
Contents lists available at ScienceDirect
Catalysis Today journal homepage: www.elsevier.com/locate/cattod
Enhancing the photoelectrochemical water splitting activity of rutile nanorods by removal of surface hydroxyl groups Shaohong Jiang, Yang Li, Xiaoli Zhang, Yongdan Li ∗ Collaborative Innovation Center of Chemical Science and Engineering (Tianjin); Tianjin Key Laboratory of Applied Catalysis Science and Technology; State Key Laboratory of Chemical Engineering (Tianjin University); School of Chemical Engineering, Tianjin University, Tianjin 300072, China
a r t i c l e
i n f o
Article history: Received 2 February 2015 Received in revised form 13 May 2015 Accepted 24 May 2015 Keywords: Rutile nanorods Surface Hydroxyl groups Photoelectrochemical water splitting Hydrogen
a b s t r a c t The effect of the surface hydroxyl (OH) groups on the photoelectrochemical (PEC) water splitting performance of rutile nanorod is examined. Based on the results of characterization, the surface OH groups are demonstrated to serve as the recombination centers for the photo-generated charges and to hinder the charge transfer at the interfaces. H2 O2 thermal and anneal treatments are applied to reduce the OH group surface coverage and both are proved effective. With reducing the relative surface OH coverage from 47 to 8%, the photocurrent density is increased from 0.419 to 0.642 mA cm−2 at 0 V (vs. Ag/AgCl), achieving 53.2% improvement. Compared with the thermal treatment, H2 O2 treatment is more effective and energy-efficient. This work highlights that the surface property of the material itself is an important factor in the design of the PEC electrodes for water splitting. © 2015 Elsevier B.V. All rights reserved.
1. Introduction TiO2 photoanode has been intensively investigated for photoelectrochemical (PEC) water splitting because of its high chemical stability, nontoxicity, and low cost [1,2]. PEC water splitting includes charge transfer steps across the semiconductor/semiconductor and semiconductor/aqueous interfaces, and surface reaction steps, which involves charge transfer and reaction. Therefore, the efficiencies of photo-generated charge separation and transfer between the interfaces are the key factors deciding the overall efficiency of the process [3–5]. Therefore, the surface modification techniques, such as co-catalysts loading and semiconductors combination, have been intensively investigated for TiO2 to enhance the photo-generated charge separation and transfer in water splitting reaction. With respect to the surface property of TiO2 itself, it has been known that the surface hydroxyl (OH) groups play an important role in the adsorption of O2 and affect the overall water splitting activity [6]. Oosawa and Grätzel and Kobayakawa et al. [7–9] investigated the effect of the surface OH density on the photocatalytic O2 evolution reaction over a rutile TiO2 powder catalyst from AgNO3 solution. Oosawa and Grätzel [7,8] found that the O2 generation activity is inversely proportional to the density of the surface OH groups. Kobayakawa et al. [9] reported that the O2
∗ Corresponding author. Tel.: +86 22 27405613; fax: +86 22 27405243. E-mail address: [email protected] (Y. Li). http://dx.doi.org/10.1016/j.cattod.2015.05.022 0920-5861/© 2015 Elsevier B.V. All rights reserved.
generation activity achieved a maximum under an optimal surface OH density. In these works, the authors ascribed the effect of the OH groups to the facilitated charge recombination on the surface OH sites. Though more direct and quantitative evidence is needed, these works show the importance of the material surface property in the design of the photocatalysts for water splitting. Recently, many works focused on one-dimensional nanostructures, because of the improvement in charge carriers separation and transport on such materials [10–13]. In 2009, Liu and Aydil [14] reported a hydrothermal growth preparation method for the oriented single-crystalline rutile TiO2 nanorods on fluorine-doped tin oxide (FTO) substrate, since then many works to enhance the PEC water splitting activity utilizing the rutile nanorods have been published, e.g. CdS and Cu2 S quantum dots-sensitization, CdS/CdSe and ln2 S3 /AgInS2 co-sensitization, as well as the branched and threedimensional hierarchical nanostructures utilizations [15–20]. It is noted that in the previous reports on the effects of the surface OH groups, rutile powder was dispersed in the solution and was utilized for the O2 evolution reaction. The understanding on the effect of surface OH groups on the PEC water splitting activity of the nanorods is meaningful. With respect to the removal of the surface OH groups, anneal treatment was practiced, which requires high temperature and is accompanied with the risks of phase transformation and sintering. H2 O2 treatment is recently used for the catalytic material surface modification. In this work, rutile TiO2 nanorods are used as the photoanode material for PEC water splitting. The role of the surface OH groups
S. Jiang et al. / Catalysis Today 259 (2016) 360–367
361
on the catalytic performance in terms of the separation and transfer of the photo-generated electrons and holes is elucidated. H2 O2 treatment (HTT) is utilized as a tool to monitor the coverage of the surface OH group. As comparison, a regular anneal treatment (AT) is also discussed. 2. Material and methods 2.1. Preparation of rutile nanorod photoanode Rutile nanorods were directly grown on a FTO substrate with a hydrothermal process [14]. Hydrochloric acid (37% by weight), titanium butoxide (AR, Tianjin Guangfu Technology Development Co., Ltd) were used as the raw materials. In a typical synthesis, two pieces of FTO substrates (3×3 cm2 ), which had been ultrasonically cleaned for 30 min in a mixed solution of deionized water, acetone and 2-propanol with volume ratios of 1:1:1, were placed at an angle against the wall of a teflon lined (100 mL volume) stainless steel autoclave with the conductive side facing down. Deionized water (30 mL) and hydrochloric acid (30 mL) was mixed and ultrasonically stirred for 5 min, and then 1 mL of titanium butoxide was added. After another 5 min of ultrasonic stirring, the solution was transferred into the autoclave. Then the autoclave was sealed and maintained at 180 ◦ C for 4 h followed by natural cooling to room temperature. The as-grown samples were rinsed with deionized water and absolute ethanol, dried in ambient air, and then annealed at 450 ◦ C for 0.5 h to enhance the contact between the rutile nanorods and FTO layer. The electrode obtained after above steps is referred to as the raw sample. 2.2. Modification of surface coverage of hydroxyl groups AT and HTT were carried out, respectively, to monitor the surface coverage of the OH groups. Samples after AT at 450 ◦ C for 1, 2, 3 and 4 h are referred to as AT-xh (AT, anneal treatment; x, treatment time). In HTT, a piece of raw sample was put into a 100 mL beaker at an angle against the wall with the conductive side facing down. After added 50 mL H2 O2 solution (30% by weight, Tianjin Guangfu Technology Development Co., Ltd) without further treatment, the beaker was placed into the water bath with the temperature maintained at 80 ◦ C for 0.5, 1, 1.5 and 2 h. Samples after HTT are referred to as HTT-xh (HTT, H2 O2 thermal treatment; x, treatment time). In addition, H2 O2 treatment was also carried out at 25 ◦ C for 0.5 and 1 h, and the samples after this are referred to as HNT-xh. 2.3. Characterization The powder X-ray diffraction (XRD) patterns were recorded with a Bruker D8 Advance X-ray Diffractometer with Cu K␣ radiation. The surface morphology was observed with a Hitachi S-4800 field-emission scanning electron microscope (SEM). The high-resolution transmission electron microscopy (HR-TEM) analysis was carried out with a JEOL JEM-2100F instrument. The UV−vis absorption spectra of the samples were acquired with an UV−vis spectrometer (PerkinElmer, Lambda 750S). The Photoluminescence (PL) spectra were obtained with Jobin Yvon Fluorolog 3-21 in 1 M NaOH solution to identify the photo-generated charges separation and utilization efficiency. The surface properties of the prepared samples were characterized by X-ray photoelectron spectroscopy (XPS, Perkinelmer, PHI1600 ESCA). 2.4. Photoelectrochemical water splitting measurements The photocurrent density (J)−potential (V) and chopped J−time (t) curves were obtained with a three-electrode setup, in which we used rutile nanorods as the working, Pt gauze as the counter,
Fig. 1. (a) X-ray diffraction patterns of the bare FTO substrate and rutile nanorods arrays with/without surface modifications, (b) UV−vis light absorption spectra of the raw sample, HTT-xh and AT-xh samples.
Ag/AgCl as the reference electrodes, and 1 M NaOH solution as the electrolyte. A 300 W Xe-lamp (Perfect Light, PLS-SXE-300) coupled with an AM 1.5G filter was applied as the light source. By changing the applied lamp current and the distance between the lamp and the rutile nanorods photoanode, the illumination intensity was adjusted to 100 mW cm−2 . The PEC data were measured using a Princeton Applied Research Versa STAT3 Potentiostat. The J−V curves were obtained with a scan rate of 5 mV s−1 , and the chopped J−t curves were obtained at an applied potential of 0 V (vs. Ag/AgCl). Electrochemical impedance spectroscopy (EIS) was measured with applying a 10 mV AC signal within a frequency range from 0.5 Hz to 100 kHz at the open-circuit potential of the PEC cells. 3. Results and discussion Fig. 1a displays the XRD patterns of the bare FTO substrate and rutile nanorods grown on FTO. The diffraction peaks of the grown samples agree well with the tetragonal rutile phase (PDF no. 88−1175), and the (1 0 1) and (0 0 2) planes can be assigned. The (0 0 2) diffraction peak appears as the most intense one,
362
S. Jiang et al. / Catalysis Today 259 (2016) 360–367
Fig. 2. SEM of the raw sample (a), HTT-2 h (b), AT-4 h (c) and HR-TEM of the raw sample.
indicating the oriented growth of rutile film. Comparing the patterns of the samples with HTT and AT at different conditions, no significant influence of HTT and AT on the crystal structure was observed. The UV−vis light absorption spectra of all the samples are shown in Fig. 1b. It is obvious that the spectra of the raw rutile nanorods film, HTT-xh and AT-xh samples are almost identical, and consistently show the onset wavelength at 413 nm (3.0 eV) agreeing with that reported in literature [1,21]. It indicates that HTT and AT have little effect on the UV−vis light absorption ability of the rutile nanorod photoanode. Fig. 2a−c shows that the rutile nanorods are tetragonal in shape. Comparing the morphology of the raw sample to that of HTT-2 h and AT-4 h that took the longest treatment time, no obvious change is noticed, which indicates that HTT and AT have no influence on the morphology of rutile nanorods. Lattice fringes of (0 0 1) and (1 1 0) planes with interplanar spacings of 0.298 and 0.322 nm, respectively, can be assigned in Fig. 2b. Plane of (0 0 1) is perpendicular to the growth direction, while that of (1 1 0) is parallel to it. This suggests that the rutile nanorods grew in the [001] direction, which is consistent with the XRD analysis and the previous results of Liu [14]. To obtain the detailed information of surface chemical states, X-ray photoelectron spectroscopy (XPS) measurements were carried out. In term of Ti 2p XPS spectra (Fig. 3a), the two main peaks, located at 458.6 and 464.4 eV, are assigned to Ti4+ 2p3/2 and Ti4+ 2p1/2 of TiO2 [22]. It is reported that the shoulder peaks at 457.5
and 463.2 eV correspond to Ti3+ 2p3/2 and Ti3+ 2p1/2 are attributed to the Ti3+ state that serves as the photo-generated charges recombination center [22,23]. To figure out the effect of HTT and AT on the Ti3+ state concentration, the relative atomic concentrations of Ti4+ 2p3/2 and Ti3+ 2p3/2 of all the samples examined are calculated, and remain identical as summarized in Table 1. This indicates that HTT and AT did not affect the Ti3+ state concentration. It is also noted here that the impurity of Cl was not detected by XPS, and that HTT and AT have little influence on the removal of the organic impurities (Supporting Information 1, SI 1).
Table 1 Surface chemical composition of all the samples measured by XPS (relative atomic concentration, %). Samples
OL
OH
Ti4+ 2p3/2
Ti3+ 2p3/2
Raw HTT-0.5 h HTT-1 h HTT-1.5 h HTT-2 h AT-1 h AT-2 h AT-3 h AT-4 h HNT-0.5 h HNT-1 h
53 65 85 92 92 59 76 84 85 56 55
47 35 15 8 8 41 24 16 15 44 45
92 92 92 92 92 92 92 92 92 92 92
8 8 8 8 8 8 8 8 8 8 8
S. Jiang et al. / Catalysis Today 259 (2016) 360–367
363
Fig. 3. X-ray photoelectron spectra of Ti 2p (a) and O 1s (b) core levels over the raw sample, HTT-xh and AT-xh.
In term of the O 1s XPS spectra (Fig. 3b), two peaks referring to two kinds of oxygen species are observed. The intense peak at about 529.97 eV is attributed to the oxygen in the TiO2 crystal lattice (OL ), while that at 531.98 eV is assigned to the surface OH [24–27]. It is obvious that the intensity of OL showed little dependence on the HTT and AT conditions, but that of the surface OH reduced along with the increase of the treatment time for each surface modification technique. This suggests that HTT and AT have little influence on the crystal structure of the rutile nanorods, which is also supported by the data of XRD, but they change the surface coverage of the OH groups. As presented in Table 1, the relative surface percentage of the OH groups of the raw sample, HTT-0.5 h, HTT-1 h, HTT-1.5 h, HTT-2 h, AT-1 h, AT-2 h, AT-3 h and AT-4 h are 47, 35, 15, 8, 8, 41, 24, 16 and 15%, respectively. With respect to each surface modification technique, the relative surface OH group coverage reduces along with the increase of the treatment time, and achieves the minimum level at 1.5 and 3 h for HTT and AT, respectively. The AT at the same temperature of HTT (80 ◦ C) was also carried out (SI 2). The result demonstrates that this temperature is not high enough to remove the surface OH groups. Therefore, under conditions of less energy intensive and shorter treatment time, HTT gives a better OH groups removal efficiency.
The PEC water splitting activity of the rutile nanorods arrays with or without HTT modification was measured with a threeelectrode electrochemical configuration in 1 M NaOH electrolyte under AM 1.5G (100 mW cm−2 ) illumination. Fig. 4a shows the J−V curves obtained by linear sweep voltammetry measurements. The photocurrent densities of the raw sample, HTT-0.5 h, HTT-1 h, HTT-1.5 h, HTT-2 h, AT-1 h, AT-2 h, AT-3 h and AT-4 h are 0.419, 0.514, 0.588, 0.642, 0.642, 0.491, 0.556, 0.587 and 0.587 mA cm−2 , respectively, at 0 V (vs. Ag/AgCl). Compared with the raw sample, HTT-0.5 h, HTT-1 h, HTT-1.5 h and HTT-2 h achieve 22.7, 40.3, 53.2 and 53.2% photocurrent density enhancement, while AT1 h, AT-2 h, AT-3 h and AT-4 h achieve 17.2, 32.7, 40.1 and 40.1% enhancement, respectively. Considering the surface OH in general, the photocurrent density increases with respect to the reduction of the relative surface OH group coverage. It indicates that the removal of surface OH groups over rutile nanorods is beneficial for PEC water splitting. The photo response and PEC water splitting stability are measured with the chopped J−t curves at 0 V (vs. Ag/AgCl) as plotted in Fig. 4b. The transient change of photocurrent density at each light-on and light-off moment demonstrates the fast photo response of the samples. A constant photocurrent density of each sample is obtained without decay in the light-on
364
S. Jiang et al. / Catalysis Today 259 (2016) 360–367
Fig. 4. PEC water splitting activity measured with a three-electrode setup in NaOH solution (1 M). (a,b) J−V curves with a scan rate of 5 mV s−1 from -0.8 to 0.5 V (vs. Ag/AgCl). (c) Chopped J−t curves at an applied potential of 0 V (vs. Ag/AgCl).
region of every cycle and shows stability through 5000s testing. This demonstrates that the J−t curves are highly reproducible and the PEC water splitting activity is stable for all the samples. PL spectrum is a facial and effective way to measure the separation and the utilization of the photo-generated electrons and holes [28,29]. The more intense the peak shows in the PL spectrum, the less separation and utilization efficiency the photo-generated charges possess. As plotted in Fig. 5, an intense peak at 427 nm appearing for all the samples is attributed to the recombination between holes trapped by surface OH groups and electrons [30]. The monotonous decrease of the intensity of this peak with respect to the reduction of the relative surface OH group coverage and the increase of the photocurrent density can be observed. As mentioned above, neither the structure, UV−vis light absorption nor Ti3+ state concentration of the rutile nanarods were affected by the HTT and AT, and no impurity was detected by XPS. The only significant difference appeared is the amount of the surface OH groups on the rutile nanorods. Therefore, all samples with PL testing have the identical bulk quality, and PL spectrum indicate the difference of charges seperation caused by the various surface OH groups coverage. Better separation and more effective utilization of the photo-generated electrons and holes can be achieved by removing the surface OH
Fig. 5. Photoluminescence spectra of the raw sample, HTT-xh and AT-xh excited at = 330 nm.
S. Jiang et al. / Catalysis Today 259 (2016) 360–367
Scheme 1. Schematic illustration of the surface OH groups on the rutile nanorods array electrode, and the recombination and transfer impediment of the photogenerated charges caused by surface OH.
groups, resulting in the enhancement of the PEC water splitting activity. The mechanism of recombination process caused by the surface OH is then verified. With respect to the photocatalytic degradation of the organic pollutant, such as methyl orange and phenol, over rutile, it is generally demonstrated and accepted that the photogenerated holes can be easily captured by the surface OH groups to produce hydroxyl radicals (Reaction 1), which possess strong oxidation ability and would trigger the subsequent degradation steps [31–34]. It is rational to assume that this reaction occurs on the rutile nanorod surface covered with OH groups. It is reported that there are three kinds of surface hydroxyl groups, Ti-(OH2 + ), Ti(OH) and Ti-(O− ), with their proportion depending on the solution pH [35–37]. In NaOH solution, surface hydroxyl is deprotonated as Ti-(O− ).
˙ Ti − O− + h+ → Ti − (O)
(1)
The hydroxyl radicals capture the photo-generated electrons, leading to the recombination (Reaction 2 and Scheme 1) and the decrease of the number of the available electrons and holes for hydrogen and oxygen evolution.
Ti − O˙ + e− → Ti − O− + hv
(2)
We agree with the suggestions of Rothenberger et al. [38] and Brown and Darwen [39] that holes are the minority carriers and should be removed rapidly from TiO2 . Co-catalysts, such as RuO2 and IrO2 , loaded on the host catalyst surface facilitates to achieve this purpose since they serve as the holes transfer catalysts. As plotted in Fig. 6, the typical EIS are presented as Nyquist plots and utilized to measure the surface charge transfer resistance. It is observed that with the reduce of the surface OH group coverage, the semicircle of the plot becomes smaller, which indicates the effective enhancement of charge transfer [40–42]. This suggests that the OH groups serve as the charge transfer obstacles over the rutile nanorod surface (Scheme 1). Taking reaction 1 into consideration, the OH groups separate the rutile nanorods and the electrolyte, and hinder the transfer of holes between these interfaces for O2 production, leading to the increase of photo-generated charges recombination probability and the ultimate reduction of PEC water splitting activity. The surface OH groups serve in two ways: (i) acting as the recombination centers for the photo-generated charges, (ii) hindering the charges transfer. These two disadvantages contribute to the degradation of the PEC performance over the rutile nanorods photoanode. In term of the recombination, the non-radiative process is the major part, and the efficiency of it is associated with
365
Fig. 6. Electrochemical impedance spectra presented as Nyquist plots of the raw sample, HTT-xh and AT-xh.
structural defects, impurities and surface trap sites [43–45]. Removal of surface OH groups may also be benefit for the inhibition of the non-radiative recombination. Further investigation on the contribution between recombination (radiative and nonradiative) and the charges transfer impediment to the degradation of PEC performance is needed in the future. To figure out the detail of the surface OH groups removal by HTT, we made a comparison experiment in which the H2 O2 treatment was done at 25 ◦ C (referred to as HNT), and compared the removal rates of the surface OH groups between HTT and HNT. The relative surface percentage of the OH groups of HNT-0.5 h and HNT-1 h are 44 and 45% (Fig. 7 and Table 1). Within the same treatment time, the surface OH group removal effect of HNT is much less efficient than that of HTT. The major difference between the two conditions is the H2 O2 decomposition rate, depending on the treatment temperature. The interaction between H2 O2 and the metal oxide surfaces was investigated in the context of nuclear technology, since H2 O2 is one of the major radiolysis products that drive corrosion of the nuclear facilities. In these precious works, it has been confirmed that the H2 O2 decomposition in the presence of metal oxide, such as TiO2 and ZrO2 , is a surface radical reaction [46,47]. The initial decomposition steps involve the molecule adsorption (Reaction 3) and the cleavage of the O O bond for hydroxyl radical production (Reaction 4), which is temperature-dependent. Therefore, more hydroxyl radicals can be produced with the increase of the treatment temperature from 25 to 80 ◦ C. In terms of the rutile nanorod surface covered with OH groups, it is proposed that the hydroxyl radicals substitute the surface OH groups forming adsorption state because of its strong bonding with the metal cation [46], resulting in the removal of the surface OH groups (Reaction 5). The subsequent H2 O2 decomposition processes are presented as the general mechanism in reactions 6−8. H2 O2 (aq) → H2 O2 (ads)
(3)
H2 O2 (ads) → 2 · OH(aq)
(4)
˙ ˙ 2 · OH (aq) + 2Ti-OH → 2 · OH (ads) /(HO) Ti O Ti (OH) + H2 O
(5)
·OH (ads) + H2 O2 (ads) → H2 O + ·OOH(ads)
(6)
·OH (ads) + ·OOH (ads) → O2 ↑ +H2 O
(7)
2 · OOH (ads) → O2 ↑ +H2 O2 (ads)
(8)
366
S. Jiang et al. / Catalysis Today 259 (2016) 360–367
Fig. 7. X-ray photoelectron spectra of (a) Ti 2p and (b) O 1s core levels over HNT-xh.
4. Conclusion With respect to the surface OH groups removal, H2 O2 treatment is more effective and energy-efficient than anneal treatment. The temperature dependence of the H2 O2 treatment indicates that the attack of hydroxyl radical produced from H2 O2 decomposition is the major reason for the removal of surface OH groups. In terms of the role of the surface OH groups on rutile nanorods for PEC water splitting, it is demonstrated that the surface OH groups serve as the recombination centers of the photo-generated electrons and holes, as well as hinder the direct transfer of the holes to the free hydroxyl ions. The PEC water splitting activity is enhanced by removing the surface OH groups. Reducing the relative surface OH coverage from 47 to 8%, the photocurrent density is enhanced by 53.2% at 0 V (vs. Ag/AgCl). To figure out the major degradation between the recombination and hindering effects over the surface OH groups, further study is needed in the future. This work highlights the crucial importance of the material surface property itself in the design of the PEC electrodes for water splitting. Acknowledgement This work has been supported in part by the Program of Introducing Talents to the University Disciplines under file number B06006, and the Program for Changjiang Scholars and Innovative Research Teams in Universities under file number IRT 0641. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.cattod.2015.05.022. References [1] [2] [3] [4] [5]
A. Fujishima, K. Honda, Nature 238 (1972) 37–38. B. O’Regan, M. Grätzel, Nature 353 (1991) 737–740. X. Chen, S. Shen, L. Guo, S.S. Mao, Chem. Rev. 110 (2010) 6503–6570. J. Yang, D. Wang, H. Han, C. Li, Acc. Chem. Res. 46 (2013) 1900–1909. X. Zhang, J. Zhao, S. Wang, H. Dai, X. Sun, Int. J. Hydrogen Energ. 39 (2014) 8238–8245.
[6] G. Munuera, V. Rives-Arnau, A. Saucedo, J. Chem. Soc. Faraday Trans. 1 Phys. Chem. Condens. Phases 75 (1979) 736–747. [7] Y. Oosawa, M. Grätzel, J. Chem. Soc. Chem. Commun. (1984) 1629–1630. [8] Y. Oosawa, M. Grätzel, J. Chem. Soc. Faraday Trans. 1 Phys. Chem. Condens. Phases 84 (1988) 197–205. [9] K. Kobayakawa, Y. Nakazawa, M. Ikeda, Y. Sato, A. Fujishima, Berichte Der Bunsengesellschaft Für Phys. Chemie. 94 (1990) 1439–1443. [10] M. Law, L.E. Greene, J.C. Johnson, R. Saykally, P. Yang, Nat. Mater. 4 (2005) 455–459. [11] T. Lindgren, H. Wang, N. Beermann, L. Vayssieres, A. Hagfeldt, S.E. Lindquist, Sol. Energ. Mater. Sol. C 71 (2002) 231–243. [12] M. Liu, N. de Leon Snapp, H. Park, Chem. Sci. 2 (2011) 80. [13] A. Wolcott, W.A. Smith, T.R. Kuykendall, Y. Zhao, J.Z. Zhang, Small 5 (2009) 104–111. [14] B. Liu, E.S. Aydil, Am. Chem. Soc. 131 (2009) 3985–3990. [15] I.S. Cho, Z. Chen, A.J. Forman, D.R. Kim, P.M. Rao, T.F. Jaramillo, Nano Lett. 11 (2011) 4978–4984. [16] K. Guo, Z. Liu, C. Zhou, J. Han, Y. Zhao, Z. Liu, Appl. Catal. B Environ. 154–155 (2014) 27–35. [17] Y.L. Lee, C.F. Chi, S.Y. Liau, Chem. Mater. 22 (2010) 922–927. [18] Z. Liu, K. Guo, J. Han, Y. Li, T. Cui, B. Wang, Small 10 (2014) 3153–3161. [19] H. Wang, Y. Bai, H. Zhang, Z. Zhang, J. Li, L. Guo, J. Phys. Chem. C 114 (2010) 16451–16455. [20] J.S. Yang, W.P. Liao, J.J. Wu, ACS Appl. Mater. Interfac. 5 (2013) 7425–7431. [21] R.I. Bickley, T. Gonzalez-Carreno, J.S. Lees, L. Palmisano, R.J.D. Tilley, J. Solid State Chem. 92 (1991) 178–190. [22] J. Zhao, J. Yao, Y. Zhang, M. Guli, L. Xiao, J. Power Sources 255 (2014) 16–23. [23] X. Zhang, H. Tian, X. Wang, G. Xue, Z. Tian, J. Zhang, Mater. Lett. 100 (2013) 51–53. [24] A. Naldoni, M. Allieta, S. Santangelo, M. Marelli, F. Fabbri, S. Cappelli, J. Am. Chem. Soc. 134 (2012) 7600–7603. [25] B. Santara, P.K. Giri, K. Imakita, M. Fujii, Nanoscale 5 (2013) 5476–5488. [26] J.C. Yu, J. Yu, H.Y. Tang, L. Zhang, J. Mater. Chem. 12 (2002) 81–85. [27] J. Zhuang, S. Weng, W. Dai, P. Liu, Q. Liu, J. Phys. Chem. C 116 (2012) 25354–25361. [28] Y. Li, X. Zhang, S. Jiang, H. Dai, X. Sun, Y. Li, Sol. Energ. Mater. Sol. C 132 (2015) 40–46. [29] X. Zhang, Y. Li, J. Zhao, S. Wang, Y. Li, H. Dai, J. Power Sources 269 (2014) 466–472. [30] S. Mathew, A.K. Prasad, T. Benoy, P.P. Rakesh, M. Hari, T.M. Libish, J. Fluoresc. 22 (2012) 1563–1569. [31] X. Fu, L.A. Clark, Q. Yang, M.A. Anderson, Environ. Sci. Technol. 30 (1996) 647–653. [32] P. Salvador, Prog. Surf. Sci. 86 (2011) 41–58. [33] A. Sclafani, L. Palmisano, J. Phys. Chem. 94 (1990) 829–832. [34] S.-J. Tsai, S. Cheng, Catal. Today 33 (1997) 227–237. [35] N. Jaffrezic-Renault, P. Pichat, A. Foissy, R. Mercier, J. Phys. Chem. 90 (1986) 2733–2738. [36] S.I. Nishimoto, B. Ohtani, H. Kajiwara, T. Kagiya, J. Chem. Soc. Faraday Trans. Phys. Chem. Condens. Phases 79 (1983) 2685–2694.
S. Jiang et al. / Catalysis Today 259 (2016) 360–367 [37] A.P. Norton, S.L. Bernasek, A.B. Bocarsly, J. Phys. Chem. 92 (1988) 6009–6016. [38] G. Rothenberger, J. Moser, M. Grätze, N. Serpone, D.K. Sharma, J. Am. Chem. Soc. 107 (1985) 8054–8059. [39] G.T. Brown, J.R. Darwent, J. Phys. Chem. 88 (1984) 4955–4959. [40] H. Zhang, X. Lv, Y. Li, Y. Wang, J. Li, ACS Nano 4 (2010) 380–386. [41] J. Meng, Z. Yu, Y. Li, Y. Li, Catal. Today 225 (2014) 136–141. [42] Z. Yu, J. Meng, Y. Li, Y. Li, Int. J. Hydrogen Energ. 38 (2013) 16649–16655.
367
[43] R. Sellappan, A. Galeckas, V. Venkatachalapathy, A.Y. Kuznetsov, D. Chakarov, Appl. Catal. B: Environ. 106 (2011) 337–342. [44] J. Buha, Thin Solid Films 545 (2013) 234–240. [45] B. Choudhury, A. Choudhury, Phys. E Low-Dimensional Syst. Nanostruct. 56 (2014) 364–371. [46] M. Jonsson, Isr. J. Chem. 54 (2014) 292–301. [47] C.M. Lousada, A.J. Johansson, T. Brinck, M. Jonsson, J. Phys. Chem. C 116 (2012) 9533–9543.