Jinghui Jiang,
Liping Tong,
Han Zhou*,
Fan Zhang,
Jian Ding,
Di Zhang and
Tongxiang Fan*
State Key Laboratory of Metal Matrix Composites, Shanghai Jiaotong University, Dongchuan Road 800, Shanghai 200240, P.R. China. E-mail: txfan@sjtu.edu.cn; hanzhou_81@sjtu.edu.cn; Fax: +86-21-3420-2749
First published on 23rd January 2015
TiO2-based photocatalysts are promising candidates for photocatalytic hydrogen evolution that utilizes solar energy due to their low cost and high stability, but their activities are usually limited by the drawbacks of TiO2. Specifically, the fluctuation of surface energy, disturbed by light irradiation, can change the relative stability of the TiO2 exposed crystal faces, which has a great influence on the photocatalytic activity. Focusing on this problem, we built a bromine containing TiO2-based photochromic self-recovery system to immobilize bromine in the photocatalyst after photocatalysis. After a series of charge migration experiments, the photocatalysis of our photochromic system was clarified, and proved that TiO2 is the final electron accepter. This conclusion further emphasizes the importance of optimizing the TiO2 surface energy to promote TiO2-based photocatalysts. As a result, a sustained effect of reducing TiO2 surface energy can be achieved by repeating the bromine adsorption to harvest stable photocatalytic hydrogen evolution under ultraviolet and visible light irradiation.
As an efficient solution of boosting TiO2 activity, non-metallic atoms have been employed to retard the increase of surface energy on exposed TiO2 facets.8,9 Halogens, a class of non-metallic elements, are efficient in reducing the surface energy of TiO2 facets by adsorbing on them.10 However, reuse of halogens in repeated photocatalytic tests is still a problem due to the weak bonding of the halogen and TiO2 facets; as a result, halogens are easily released into the environment during the after-treatment. Herein, for the purpose of harvesting a sustained effect of reducing the TiO2 surface energy, an efficient solution of immobilizing halogens in the photocatalyst after photocatalysis is explored to maintain photocatalytic activity, save costs, and avoid environmental pollution.
Based on the above assertion, acetylacetone is utilized as a ligand to extract Ti3+ ions from titanium(III) chloride solution to prepare a titanium precursor without chlorine ions (the detailed process is listed in the Experimental section). Due to the absence of chlorine ions in the titanium precursor, silver bromide (AgBr) and cupric oxide (CuO) are employed to build a photochromic self-recovery system. The light yellow AgBr, decomposed by light irradiation, can regenerate again from black silver (Ag0) and dissociative bromine molecule (Br2) in the dark by CuO catalysis to achieve the cyclic process of color change – the photochromic effect.11,12 Contributing to this characteristic, a well distributed TiO2-based photochromic nano-system (TPN) can be synthesized by merging this photochromic system into the titanium precursor via a chemical synthesis step and one sintering process. Fortunately, AgBr can recover from the photolysis to achieve the goal of immobilization of Br− in TPN after photocatalysis. Meanwhile, Ag0, from AgBr photolysis, and CuO are also introduced to modify the structure of TiO2, in this way, to facilitate photon capture and charge separation.13–15 To the best of our knowledge, this strategy is novel, simple, and cost efficient in maintaining stable and highly efficient photocatalytic hydrogen evolution while avoiding environmental pollution due to the environment unfriendliness of Br2. Moreover, it also overcomes the significant shortcomings, such as non-uniform dispersion, weak component bonds, and limited interface, of the deposition–precipitation method – the strategy commonly employed in previous research.16
The preparation process for TiO2/AgBr, TiO2/CuxO, and TiO2 is the same as TPN, except for the addition of silver nitrate, potassium bromide, or copper nitrate trihydrate: specifically, adding silver nitrate and potassium bromide to form TiO2/AgBr, adding copper nitrate trihydrate to form TiO2/CuxO, and adding none of this to form TiO2.
To investigate in detail the shape and exposed crystal faces of TPN, transmission electron microscopy (TEM) and high-resolution transmission electron microscopy (HRTEM) images were collected. The as-synthesized TPN (see Fig. 1c) looks like irregular flakes with diameters of 20 to 40 nm. Fig. 1d is a typical HRTEM image for the as-synthesized TPN, which reveals the crystal lattice and the orientation of the lattice planes with respect to the crystal faces. Fig. 1e–g shows the fast Fourier transform (FFT) patterns taken from zones 1, 2, and 3 in d, respectively. We can calculate from Fig. 1e that zone 1 in Fig. 1d is anatase, and along the [010] direction presents (101) and (10) crystallographic planes with a lattice spacing of 0.351 nm. The indicated angle in the FFT image (see Fig. 1e) is 41.3°, which is identical to the theoretical values of the angle between {101} faces. This information confirms that anatase nanocrystals in TPN mainly expose {101} faces to offer photo-reduction sites for hydrogen evolution.5 Similarly, we can further confirm zone 2 and zone 3 in Fig. 1d are AgBr and CuO crystals with exposed crystallographic planes of (1
1) and (
11), respectively.
Since TiO2, CuxO, and AgBr are coexisting in TPN, ESR at 100 K was employed to confirm that TiO2 is the final electron acceptor in TPN during photocatalysis. This is the premise of studying the effectiveness of Br2 on TiO2 photocatalytic activity by excluding the possibility of CuxO or AgBr photocatalytic hydrogen evolution. Compared to samples without light irradiation (see Fig. S2a†), the enhanced paramagnetic signals of the irradiated samples (see Fig. 2a) indicate a series of charge transfer processes took place in TPN during light irradiation. The spectra of irradiated TiO2 and TiO2/AgBr (see spectra 1 and 2 in Fig. 2a) were collected to investigate the influence of AgBr on charge transfer. Obviously, the signal at g = 1.992, the characteristic signal of Ti3+,22 in spectrum TiO2/AgBr is largely enhanced, indicating that Ag0, from AgBr photolysis (see Fig. S3†), can efficiently boost Ti3+ trapping electrons; specifically, Ag0 not only can provide photon capture sites to trap photoelectrons, but it can form a Schottky barrier with TiO2 at their interface to facilitate electron transfer from Ag0 to TiO2 due to their work function (WF), (φAg = 4.25 eV, φTiO2 = 4.6 eV).23,24 Based on electron trapping, the marked signals in spectrum TiO2 with relative intensities of 1:
2
:
1 can be assigned to methanol radical [ĊH2O(H)], due to oxidation,25 which disappears in spectrum TiO2/AgBr. The most reasonable explanation is holes (h+) are trapped by Br− ions to generate Br2 molecules;26 thus, this signal will not arise in spectrum TiO2/AgBr before all the Br− ions are oxidized. This control experiment reveals that Ag0 can help TiO2 to harvest photoelectrons, and TiO2 is the final electron acceptor in TiO2/AgBr.
As for the influence of CuxO (x = 1, 2), some new signals emerge in spectrum TiO2/CuxO (see spectrum 3 in Fig. 2a), except the signal at g = 1.992 and the marked signals. Obviously, the signal at g = 1.992 in spectrum TiO2/CuxO is even stronger than that in spectrum TiO2/AgBr, indicating CuxO is more efficient than AgBr at facilitating Ti3+ trapping electrons. This strengthening should be contributed by the p–n junctions and conduction-band edges of TiO2 and CuxO (VCB,TiO2 = −0.52 eV, VCB,CuO = −0.3 eV, VCB,Cu2O = −1.13 eV),27–29 which facilitate the migration of photoelectrons from Cu2O to TiO2 when all CuO particles are reduced to Cu2O30 (see spectrum 2 in Fig. S4†); as a result, more electrons will enrich in TiO2 due to the increase of visible light absorption by Cu2O doping. Furthermore, axially symmetric signals with g‖ = 2.330, A‖ ≈ 100 G hyperfine splitting are assigned to Cu2+ ions at the cationic substitution sites of TiO2, which are produced by some Cu2+ ions replacing surface Ti4+ sites in thermal treatment at 773 K.31 In addition to the above signals, a new signal at g = 1.892 for the Ti(H2O)63+ complex,32 formed by rapid freezing at 100 K, appears in spectrum TiO2/CuxO, indicating CuxO can facilitate Ti3+ to absorb water molecules. Lastly, the marked signals in spectrum TiO2/CuxO are the same as in spectrum TiO2 due to there being no Br− ions to consume the holes. Therefore, TiO2 is the final electron acceptor in TiO2/CuxO during photocatalysis, as nearly all photoelectrons, harvested by Cu2O, will transfer to TiO2.
Based on the above analyses, the changes in TPN (see spectrum 4 in Fig. 2a) become easier to understand. When Ag0, Cu2O, and TiO2 form a well-distributed system after photolysis, Ag0 acts as a photon capture site to trap photoelectrons and ohmic contact of Ag0 with Cu2O or TiO2 due to WF (φCu2O ≈ 4.84 eV)33 facilitates photoelectrons transferring from Ag0 to Cu2O or TiO2. Additionally, well-distributed p–n junctions at the interface of Cu2O and TiO2 facilitate the transportation of photoelectrons to TiO2. As a result, Ti3+ trapping electrons is enhanced to the largest extent by the assistant of Ag0 and Cu2O, which can be verified by the strongest signal at g = 1.992 in spectrum TPN. Similarly, the appearance of the Ti(H2O)63+ complex in spectrum TPN is in agreement with the discussion in spectrum TiO2/CuxO. Lastly, a broadening signal in spectrum TPN, around g = 2.240, compared to spectrum TiO2/CuxO, is caused by the increase of the paramagnetic signal concentration as AgBr and CuxO are added to form TPN. According to the above analyses, TiO2 is the final electron acceptor in TPN during photocatalysis, and the photochromic system is efficient in helping TiO2 to harvest light and to contribute an ideal quantum yield, approximately 14.3%, at 340 nm wavelength (see Table S3†).
The corresponding photocatalytic hydrogen evolutions (see Fig. 2b) were further conducted to verify the above ESR analyses. Apparently, the increase of photocatalytic hydrogen evolution for TiO2, TiO2/AgBr, TiO2/CuxO, and TPN coincides with their signal intensity of Ti3+ at g = 1.992 in Fig. 2a, proving TiO2 is the final electron acceptor in TPN and Ti3+ in anatase is the main active site for photocatalytic hydrogen evolution. Following that, another platinum (Pt) photodeposition experiment was further conducted (see Fig. S5†), where Pt selectively deposited on some crystal particles in TPN, indicating those particles enrich electrons. HRTEM analyses reveal those crystals, in contact with Pt, are TiO2 (see Fig. S5b–d†), which further firmly confirms that TiO2 is the final electron acceptor in TPN. Therefore, our strategy of promoting the photocatalytic hydrogen evolution by building an AgBr/CuxO self-recovery system to repeatedly optimize the surface energy of TiO2 in TPN is reasonable.
Following that, five consecutive photocatalytic hydrogen evolution and dark self-recovery cycles (see Fig. 3a) were conducted to verify the effectiveness of AgBr self-recovery on photocatalytic hydrogen evolution. The yield of recovered TPN is high from run 1 to run 4 with some reduction in run 5, but a gradual decrease occurs for unrecovered TPN indicating the necessity of AgBr self-recovery in maintaining the photocatalytic activity. To reveal the difference between recovered and unrecovered TPN, XRD and XPS tests were conducted to confirm their components and valences. Specifically, the component (1 to 5 in Fig. 3b) and valences (1 and 3 in Fig. 3c and d) of AgBr in recovered TPN are nearly the same as in pristine TPN, revealing AgBr can also self-recover in the TiO2 matrix. Contrarily, Ag0 arises in unrecovered TPN after a photocatalytic process (see 6 in Fig. 3b and 2 in Fig. 3c), and the signal of Br− 3d5/2 at 68.6 eV (see 2 in Fig. 3d) also disappears in unrecovered TPN indicating the missing of Br− in unrecovered TPN. The above analyses reveal that all AgBr in TPN are decomposed to Ag0 and Br2 during photocatalysis, thus the dissociative Br2 has the opportunity to adsorb on TiO2. As shown in Fig. S4,† the specie of Cu2p exists in pristine, unrecovered, and recovered TPN, indicating the changes of CuxO during photocatalysis can be ignored. Comprehensive consideration of the above analyses, the difference between recovered and unrecovered TPN is the existence of Br−, therefore, the difference of photocatalytic hydrogen evolution between recovered and unrecovered TPN shown in Fig. 3a can only be aroused by Br2. Since TiO2 is the final electron acceptor in TPN, the stable performance of recovered TPN can be attributed to the adsorption of Br2 on TiO2 to modify its surface energy.
To verify the above conclusion, the first-principles density functional theory (DFT) was used to calculate the chemisorption of Br2 on TiO2 (101) surface, since the anatase nanocrystals in TPN have been proven to expose {101} faces to offer the photo-reduction sites in Fig. 1d. The calculations performed in this study were done using the CASTEP of Materials studio. Specifically, models of Br2 adsorption on a layer of atom in (2 × 2 × 1) lattice cell of TiO2 (101) facet at different sites (see Fig. 4a–e) were built to investigate the influence of Br2 adsorption on chemisorption energy (the detailed calculation is listed in Table S4†). According to the calculation, all values of chemisorption energy (ΔEchem) (see Fig. 4f) are negative, indicating Br2 adsorption on TiO2 (101) facet is able to reduce its surface energy and to pose a positive influence on photocatalytic hydrogen evolution. Herein, our strategy of promoting the performance of TiO2-based photocatalytic hydrogen evolution by building an AgBr self-recovery system to repeatedly optimize the surface energy of TiO2 is successful and effective.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra15416a |
This journal is © The Royal Society of Chemistry 2015 |