Zhao Gao1, Hanpei Yang1, Hongyu Zhu1, Runqiang Guo1, Junmin Wu1. 1. Key Laboratory of Integrated Regulation and Resource Development on Shallow Lakes, Ministry of Education, College of Environment, Hohai University Nanjing 210098 China yanghanpei@hhu.edu.cn +86 25 83786090 +86 25 84968465.
Over the past few decades, titanium dioxide (TiO2) has been widely studied in pollution control.[1] However, it usually shows inertness under visible light irradiation and low quantum yield in light energy utilization.[2] Recent strategies focused on grafting photosensitizers or semiconductors onto TiO2 for expanding their ranges of light absorbance and inhibiting the recombination of photogenerated electron–hole pairs.[3,4]For an effective photosensitizer of TiO2, two important criteria are required:[5] the photoactive compound should have a high extinction coefficient in the visible region, and be capable of being adsorbed on the TiO2 surface via physical/chemical interaction. In constructing heterojunctions, the band gap of the semiconductor used to modify TiO2 should be narrow, and its conduction and valence band positions should be matched with that of TiO2, respectively.[4] On these accounts, metal phthalocyanines (MPcs, M = Fe, Co) are benign candidates for the modification of TiO2. More interestingly, the M–N4 structure in MPcs can increase the O–O length of oxygen, which would play an important role in promoting the production of superoxide radical (·O2−) from O2.[6] Many efforts have been devoted to coupling of MPcs with TiO2.[3,7,8] However, the recognition on the versatility of MPcs in photocatalysis is insufficient, especially, the function of MPcs in activating oxygen in the process of degrading organic pollutants.In this report, cobalt phthalocyanine sulfate (CoPcS) was composited with TiO2, and the experiments on MB degradation over CoPcS/TiO2 were conducted. The multiple roles of CoPcS in composite were investigated and synergy in photosensitization, charge separation and oxygen activation was proposed.
Experimental
Synthesis of samples
All the reagents used in this experiment were received without further purification. The CoPcS/TiO2 composite was fabricated via a hydrothermal route. In a typical preparation, 60 mg CoPcS (optical) and 20 ml absolute ethyl alcohol were mixed and ultrasounded for 30 min to get a homogeneous turbid liquid. Subsequently, another 16 ml absolute ethyl alcohol, 3.2 ml acetic acid and 10 ml Ti(C4H9O)4 were added into the mixture, followed by dropwise addition of 2 ml deionized water under vigorous stirring for 1 h. Then, the compound was loaded into a 100 ml stainless steel autoclave, sealed and moved into an oven and kept at 180 °C for 10 h. After cooling the autoclave to room temperature, the precipitate was washed with ethanol and deionized water thrice, and then dried at 80 °C for 24 h. Finally, the solid was annealed at 300 °C for 2 h in purity N2. For comparison, TiO2 were prepared under same procedure without adding of CoPcS.
Characterizations and measurements
X-ray diffraction (XRD) analysis were performed on a Shimadzu-3A diffractometer at 40 kV and 30 mA with Cu Kα radiation (λ = 0.15418 nm). The morphologies were examined by transmission electron microscopy (TEM, JEM-2100CX, JEOL). Infrared spectra (FT-IR) were acquired with an 8400S spectrometer (Shimadzu) in the transmission mode. X-ray photoelectron spectra (XPS) were obtained by a PHI 5000 Versa Probe spectrometer (ULVAC-PHI) operated at a voltage of 13 kV and an emission current of 28 mA using Al Kα as exciting source (1486.6 eV). The binding energies were referenced to C 1s at 284.5 eV. The UV-vis absorption spectra of samples were obtained from a Shimadzu UV-3600 spectrophotometer equipped with an integrating sphere using BaSO4 as reference. Photoluminescence (PL) spectra were recorded on F-7000 fluorescence spectrophotometer (Hitachi) with a laser excitation of 420 nm. Electron paramagnetic resonance (EPR) signals of paramagnetic species spin-trapped with DMPO were recorded at ambient temperature (298 K) with a Brucker EPR 300E spectrometer, the irradiation source (λ = 532 nm) was a Quanta-Ray Nd:YAG (10 pluses per second) laser system.
Photocatalytic oxidation experiments
The visible-light-driven photocatalytic activity of the as-prepared samples was monitored from the results of the degradation of MB. For each photocatalytic activity measurements, 10 mg of as-prepared catalysts were dispersed into 100 ml of MB solution initialized at 5 mg L−1. The light comes from a 300 W xenon lamp (CEL-HXF-300, Education Au-light Co., Ltd., Beijing, China) equipped with a UV cutoff filter (λ ≥ 400 nm). The photocatalytic reactions took place in the reactor connected to a water bath to main the solution at about 25 °C and the reaction aqueous slurries were magnetic stirred and bubbled with air at a flow rate of 40 ml min−1. The suspension was stirred in the dark for 1 h to obtain adsorption equilibrium of MB before illumination. During the photo-reaction, samples were collected at selected time intervals. The catalyst powders were removed by filtration and the residual concentration of MB was determined by the spectrophotometer. Quenching experiments were conducted under same conditions except the existence of each scavenger in 10 mM of ethylenediamine tetraacetic acid disodium (EDTA-Na2, for ·O2−), tert-butyl alcohol (tBA, for ·OH) and p-benzoquinone (pBQ, for h+).
Results and discussion
Morphology and structure
XRD and TEM analysis
Fig. 1A shows the XRD patterns of the as-prepared samples. The spectrum of bare TiO2 and CoPcS/TiO2 show the typical peaks of anatase phase (JCPDS no. 21-1272), while the diffraction peaks of CoPcS were not observed on the XRD pattern of CoPcS/TiO2 probably due to the low loading mass or small size of loaded CoPcS.[9] The average crystallite sizes of pure TiO2 and CoPcS/TiO2 were calculated to be 9.9 and 9.2 nm, respectively, based on the Scherrer formula. TEM observations of CoPcS/TiO2 (Fig. 1B) indicate an intimate coating of CoPcS on TiO2, and particle sizes roughly matched to that from XRD. As shown in Fig. 1C, the HRTEM image of CoPcS/TiO2 displays two types of clear lattice fringes, one set of the fringe spacing (d) was ca. 0.35 nm, corresponding to the (101) plane of the anatase crystal structure of TiO2;[10] another set of stacking feature (d ≈ 0.252 nm) corresponds to the CoPcS.[11] The result agrees well with that from the XRD analysis.
Fig. 1
XRD patterns of samples (A), TEM (B) (inset is that of pure TiO2) and HRTEM (C) image of CoPcS/TiO2.
FTIR analysis
The surface structures of resultant samples were revealed by FT-IR spectra as shown in Fig. 2A. The spectrum of pure TiO2 shows the Ti–O–Ti at 539 cm−1, Ti–O–H at 1654 and 3468 cm−1.[12,13] The spectrum recorded on CoPcS/TiO2 shows distinct difference from what on pure TiO2 with C–C at 1404,[14] CC and CN at 1638 cm−1.[15,16] The peak centered at 917, 1040 and 1232 cm−1 is attributed to Co–N,[17] C–N[14] and S–O[18] in CoPcS, sequentially. The broad peak at 608 cm−1 is induced by Ti–O–Ti and Ti–O–S. This is a strong evidence of covalent attaching of CoPcS on TiO2. The linkage between CoPcS and TiO2 is proposed as Fig. 2B.
Fig. 2
FT-IR spectra on TiO2 and CoPcS/TiO2 (A) and the possible linkage between CoPcS and TiO2 (B).
XPS analysis
The surface structure of CoPcS/TiO2 was confirmed by XPS. In Fig. 3A, the peaks of C 1s in CoPcS/TiO2 can be deconvoluted into four lines peaked at 283.7, 284.5, 285.5 and 287.2 eV, corresponding to CC, C–C, C–N and CN in CoPcS, respectively.[19-22] The O 1s (Fig. 3B) composed of three peaks, the deconvoluted peak observed at 529.3 eV corresponds to the Ti–O–Ti in TiO2. The peak at a binding energy of 530.4 eV is attributed to Ti–O–S,[23] indicating a covalent linkage between CoPcS and TiO2. An obvious component with the binding energy at 532.4 eV can be assigned to the oxygen (*O2) coordinated by CoPcS.[24]
Fig. 3
High-resolution XPS spectra of C 1s (A) and O 1s (B) in CoPcS/TiO2.
Photocatalytic activity of samples
As shown in Fig. 4A, the removal of MB by direct photolysis or photocatalytic degradation on CoPcS was observed negligible. Pure TiO2 exhibited nearly 28.3% of MB degradation mainly attributed to their visible-light-driven activity under self-photosensitization of MB.[25] The CoPcS/TiO2 exhibited superior performance on the MB degradation, with the degradation rate of 88% and the pseudo-first-order rate constant of 0.0091 min−1 (almost 6.2 times of that on pure TiO2). Remarkably, obvious decrease in MB degradation was observed on CoPcS/TiO2 under anaerobic conditions (by bubbling N2), suggesting that O2 was crucial in the reaction.
Fig. 4
Photocatalytic degradation of MB on samples (A) and pseudo-first order fitting of the photocatalytic data (B), (C0 in (A and B) represent the actual concentrations of MB after their adsorption–desorption equilibrium in the dark).
Versatility of CoPcS in CoPcS/TiO2 for MB degradation
Photosensitization
As depicted in Fig. 5A, the absorption spectrum recorded on TiO2 exhibited a typical behavior of a wide-band-gap oxide semiconductor, with no absorption in visible region. However, the CoPcS/TiO2 exhibited strong absorption of light in whole wavelength region. Moreover, the spectrum showed an obvious red-shift of absorption edge to approximately 445 nm, and typical peaks of Q band from dimer and monomer CoPcS at 603 and 669 nm−1[17] resulted from the excitation from their HOMO to the LUMO.[26,27] Compared to the regular CoPcS, the peaks in Q band of CoPcS/TiO2 exhibited red and blue shifts slightly, suggesting the electronic coupling between CoPcS and TiO2 due to the Ti–O–S linkage indicated by IR and XPS.[28,29] The energy band gaps from the UV-vis DRS spectra were deduced from the Tauc plot using the Kubelka–Munk theory, and the result was shown as Fig. 5B. The band gap energy of TiO2 was determined as ∼3.2 eV, while that of CoPcS/TiO2 was calculated to be ∼2.7 eV, which matched well with the absorption edge at 445 nm.
Fig. 5
UV-vis diffuse reflectance absorption spectra (A) and their Tauc plot (B) of TiO2 and CoPcS/TiO2.
Under visible light irradiation of CoPcS/TiO2, the singlet excited state (S1) of CoPcS would typically generated from the ground state (S0) and then transformed to triplet excited state (T1) through innersystem crossing.[30] The redox potential of S0, S1 and T1 of CoPcS are around 0.46, −1.35 and −0.75 eV (vs. NHE), respectively.[31,32] The generation of S1 is normally negligible due to their short lifetime (ns),[33] but the excited CoPcS in T1 (ms) can inject charges into the conduction band of TiO2, generating cation radicals of CoPcS (CoPcS+˙). The CoPcS+˙ can participate directly in the degradation of MB[33] and contributes to the enhanced activity of CoPcS/TiO2 showed by Fig. 4. Herein, the photosensitization of TiO2 by CoPcS can be expressed as follows:
Charge separation
It is well accepted that CoPcS is a typical narrow-band-gap semiconductor with its Eg of about 2.1 eV.[34] Coupling TiO2 with CoPcS also contributed to the absorption of visible light on CoPcS/TiO2. As indicated in Fig. 5, the CoPcS/TiO2 showed significant light adsorption in 445–595 nm, which was consistent with the band gap of CoPcS. In addition, the formed heterojunction between CoPcS and TiO2 played an important role in the separation of photogenerated electron–hole pairs. The conduction band edges of TiO2 and CoPcS are −0.5 and −1.05 eV, respectively.[35-38] The photogenerated electrons were able to transfer from the conduction band of CoPcS to that of TiO2, leaving holes on the valence band of CoPcS. In this way, the photogenerated electron–hole pairs on CoPcS got separated.Charge separation on the heterojunction was confirmed by PL measurement. As demonstrated by Fig. 6, the remarkable decrease in PL intensity demonstrated that deposition of CoPcS onto TiO2 decreased the carrier recombination rate and improves the separation efficiency of photogenerated electrons and holes, which was favorable to the degradation of MB.[39]
Fig. 6
PL spectra of TiO2 and CoPcS/TiO2.
Oxygen activation
As identified by the quenching experiments illustrated in Fig. 7, ·O2−, ·OH and h+ all played significant roles in proceeding MB degradation, especially the ·O2−, which is generated predominantly through the trapping of photo-excited electrons by dissolved molecular oxygen. For the redox potential of holes on the valence band of CoPcS is negative than E (H2O/·OH), we deduce that ·OH is generated from ·O2−.[40]
Fig. 7
Degradation rates of MB over CoPcS/TiO2 in solutions with and without scavengers.
The presence of ·O2− and ·OH radicals was further confirmed by the electron spin response (ESR) experiments of CoPcS/TiO2 with 5,5-dimethyl-1-pyrroline (DMPO) as a scavenger in a methanol and an aqueous solution. The four characteristic peaks of DMPO-·O2−[41] (Fig. 8A) and the system signature (1 : 2 : 2 : 1 signals) of DMPO–·OH radical adducts[42] (Fig. 8B) were both observed. In contrast, no DMPO-·O2− and DMPO-·OH signals emerged for bare TiO2 dispersion.
Fig. 8
DMPO spin-trapping ESR spectra recorded with as-prepared samples in (A) methanol dispersion (for DMPO–·O2−) and (B) aqueous dispersion (for DMPO–·OH) under visible light irradiation.
According to the above results, we consider that oxygen in the reaction is activated by the Co–N4 structure in CoPcS. The electronic configuration of 3d orbital of free Co2+ (in spherical field) is diagrammatically presented as Fig. 10 (a). In a square-planar crystal field offered by CoPcS in Fig. 2B, the degenerate energy level of 3d-orbitals split into four levels as sketched as (b).[43] Coordination of dioxygen (as a fifth ligand[44]) to Co2+ surrounded by the macrocyclic ligand as CoPcS cause a further rearranging of energy into two levels with eg and t2g symmetry as (c) in Fig. 9.[45] However, this octahedral symmetric configuration in a non-liner molecular is instable due to the non-full occupation in 3d orbital of Co2+,[46] the configuration will be distorted as (d) by a Jahn–Teller effect.[47]
Fig. 10
A schematic diagram of the synergetic mechanism in MB degradation on CoPcS/TiO2.
Fig. 9
Sketch of the energy splitting of Co2+ in different crystal field.
In such a configuration, most of the interpretations of experimental and theoretical investigation coincided in the conclusion that the 3d orbital is half filled.[48] A σ-rich orbital of O2 donates electron density to 3d of Co2+, forming a σ-type bond, while a π interaction is produced between the dπ (d, d) orbitals and π* orbitals of dioxygen, with charge transfer from metal to O2.[49] This electrons rearranging get oxygen activated and increase the O–O bond length from the usual 1.21 to ∼1.30 Å.[6] According to literature, the redox potential of oxygen in ground state E (3O2/·O2−) is around −0.048 eV.[50] However, with the activation, the potential value can increase to ∼0.77 eV,[6] which is more positive than that of ECB in TiO2.Based on the above results, the synergy of photosensitization, charge separation and oxygen activation on CoPcS/TiO2 was proposed as Fig. 10. The electrons generated by photosensitization and charge separation on the conduction band of TiO2 can be more easily trapped by the activated oxygen (*O2), derivating more ·O2− species participating in degrading MB. Some of the ·O2− reacts with H+, followed by producing ·OH of a redox potential of 2.4 eV. The redox potentials of generated CoPcS+˙ and holes were 1.2 and 1.05 eV, respectively.[37,38] Thus, the MB was oxidized by ·OH, CoPcS+˙ and holes. By this way, the photocatalytic activity of CoPcS/TiO2 in degrading MB was enhanced.
Conclusion
In this work, TiO2 was composited with CoPcS via the Ti–O–S linkage. The photosensitization of TiO2 by CoPcS and charge separation on the heterojunction were promoted. At the same time, the oxygen was activated by CoPcS. Due to the versatility of CoPcS on TiO2, the degradation rate of MB over CoPcS/TiO2 reached 88% under visible light in 4 h. This synergy is of great potential for design of high-photoreactive catalysts using CoPcS as a component.
Authors: N Sundaraganesan; S Kalaichelvan; C Meganathan; B Dominic Joshua; J Cornard Journal: Spectrochim Acta A Mol Biomol Spectrosc Date: 2008-02-14 Impact factor: 4.098
Authors: T Kroll; V Yu Aristov; O V Molodtsova; Yu A Ossipyan; D V Vyalikh; B Büchner; M Knupfer Journal: J Phys Chem A Date: 2009-08-06 Impact factor: 2.781