SEARCH

SEARCH BY CITATION

Keywords:

  • doping;
  • iron;
  • photochemistry;
  • nanostructures;
  • titanium

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and Discussion
  5. Conclusions
  6. Experimental Section
  7. Acknowledgements
  8. Supporting Information

Despite sustained effort over the years, additional investigations into the relationship between doping elements and the photocatalytic properties of TiO2 are needed. In this work, Fe-group(Fe, Co, Ni)-doped anatase TiO2 nanosheets with exposed {0 0 1} facets are prepared by using a facile solvothermal synthetic route. A change in the type and concentration of the doping element consequently leads to a change in the size and percentage of {0 0 1} facets in the TiO2 nanosheets, and finally causes a significant difference in the photocatalytic activities of the as-prepared TiO2 samples. In this case, the performance of doped TiO2, which serves as the photocatalyst, is investigated by directly assessing its effect on the photodegradation of azo dyes (methylene blue) under UV and visible light irradiation, respectively. Importantly, for the first time these studies reveal that the order of the activities of the doped TiO2 is Ni>Co>Fe, and the optimal value of the percentage of the doping ions to Ti ions is 0.75 %. Electrochemical impedance spectra, photocurrent, and photoluminescence measurements indicate that the charge separation and transportation rate is mainly influenced by the type and concentration of doping ions, which determines the photodegradation activities of the TiO2 host.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and Discussion
  5. Conclusions
  6. Experimental Section
  7. Acknowledgements
  8. Supporting Information

The development of an efficient and stable photocatalyst for remediating global air and water pollution is an important research topic. Since Frank and Bard first examined the possibilities of using TiO2 powder to decompose cyanide in water,1 there has been an increasing interest in environmental applications. With its chemical stability, low cost, low toxicity, and high efficiency in removing pollutants from water and air, TiO2 is one of the most studied photocatalytic materials today. The photocatalytic process at the surface of TiO2 usually involves the absorption of photons with energies that are larger than the band gap of TiO2. Excited electrons are transported from the valence band to the conduction band, which creates electron–hole pairs. These charge carriers migrate to the surface of TiO2 and react with, and decompose the chemicals adsorbed on the surface.24 The photocatalytic activity of TiO2 is largely determined by: 1) the light absorption properties, such as the light absorption spectrum and coefficient; 2) the electron–hole recombination rate; and 3) the available specific surface area. However, the full potential of TiO2 has not been achieved for two reasons. First, the wide band gap (3.2 eV for anatase phase) indicates that TiO2 is only active in the UV region, which accounts for less than 5 % of the total energy in the solar spectrum. Second, the efficiency of charge carrier transport is low, which owes to the rapid recombination rate between photoelectrons and holes. Therefore, addressing the effective utilization of visible light and photogenerated charge carriers is a crucial issue for increasing the usefulness of TiO2 as a photocatalyst.

As far as we know, the optical response of any material is mostly determined by its fundamental electronic structure. The electronic properties of a material are closely related to its chemical composition (this includes the elemental and chemical nature of the bonds between the atoms or ions), its atomic arrangement, and its physical dimension (confinement of carriers) for nanometer-sized materials. Doping is a strategy that can alter the chemical composition of TiO2. Specifically, the material’s optical properties can be altered by replacing the metal (Ti) or the nonmetal (oxygen) component with another element.5, 6 It is necessary to maintain the integrity of the crystal structure of the photocatalytic host material to produce the available changes in the electronic structure. It appears to be more difficult to replace the O2− anion in TiO2 with other anions than it is to substitute the Ti4+ cation with other transition metals, owing to differences in charge states and ionic radii.7 Consequently, several metals have been doped into TiO2 nanomaterials, such as Ti, V, Cr, Mn, Fe, Co, Ni, Cu, Nb, Mo, Sn, Er, Au, and Bi.815 Choi et al.16 systematically studied the photocatalytic activity of quantum-sized TiO2 doped with 21 metal ions by the sol–gel method and found that the presence of metal ion dopants significantly influenced the photoreactivity, charge carrier recombination rates, and interfacial electron-transfer rates. Anpo et al.17 prepared V, Cr, and several transition metals for doping TiO2 with an ion-implantation method, and showed that implanted ions shifted the absorption band to the visible light region, thereby increasing the photoreactivity of TiO2 in the visible region. In contrast, Herrman et al.18 and Borgarello et al.19 indicated that Cr-doped TiO2 synthesized by a chemical method decreased the photoreactivity under both visible and UV light illumination. A theoretical analysis through a computer simulation is expected to clarify the impurity-doping effects on the electronic structure of the host material. The electronic structures of V, Cr, Mn, Fe, Co, and Ni-doped rutile TiO2 were analyzed by Umebayashi et al.20 They reported that if TiO2 was doped with the above ions, an electron-occupied level formed and the electrons were localized around each dopant. Li et al.21 found that the substituted Nd3+ ion introduced electron states into the band gap of TiO2 to form the new lowest unoccupied molecular orbital and thus reduced the band gap of TiO2. It is now commonly recognized that the structure-performance relationships of photocatalysts strongly depend on the preparation methods and their size, shape, and surface structure. Even though the effects of metal-ion doping on the activity have been investigated frequently, it remains difficult to make direct comparisons and unifying conclusions because a wide series of experimental conditions have been used in sample preparations and the determination of photoreactivity. Furthermore, until now, the synthesis of well-defined Fe-group ion-doped anatase TiO2 nanosheets with exposed {0 0 1} facets has never been reported, and there are few reports of a direct correlation between the photocatalytic reactivity and similar metal ion dopants in anatase TiO2 nanosheets.

To better investigate the doping effects, it is first necessary to synthesize well-defined Fe-group ion-doped anatase TiO2 nanosheets with exposed {0 0 1} facets. In the present study, Fe3+, Co2+, and Ni2+ ions have been chosen as dopants, because they all have similar physicochemical properties; in addition, they cause TiO2 to have a visible light photoresponse.10, 16, 22 The size and percentage of {0 0 1} facets in TiO2 nanosheets can be tuned by varying the concentration of metal nitrates that are added. The photocatalytic properties of doped TiO2 nanosheets are investigated by the degradation of methylene blue (MB) in the presence of UV light (λ<400 nm) and visible light (λ>420 nm), respectively. A series test proves that changing the type and concentration of the dopant causes a significant variation in the charge-pair recombination rate and optical response of the as-synthesized samples. Notably, the difference in the photoreactivity can be seen, the order of the enhancement in the photocatalytic activity is Ni>Co>Fe. We consider that the type of doping ions directly influences the electronic states of doped TiO2, which contributes to the charge-carrier recombination rate. Furthermore, we believe that the dual role of doping ions (positive trapping sites or negative recombination centers) can be converted depending on the doping level, and appears to be the most important factor in the photocatalytic properties of doped TiO2.

Results and Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and Discussion
  5. Conclusions
  6. Experimental Section
  7. Acknowledgements
  8. Supporting Information

Synthesis and characterization of doped TiO2 nanosheets

According to first-principle calculations by Yang et al.,23 fluoride ions can markedly reduce the surface energy of the {0 0 1} facets to a level lower than that of the {1 0 1} facets. In this work, doped TiO2 nanosheets could be facilely synthesized by an improved fluoride-mediated self-growing procedure. Meanwhile, to eliminate the interference of the anion, we select nitrate as the anion for the whole procedure. All of the samples are denoted as M-X, in which M refers to Fe, Co, and Ni, and X indicates the M/Ti atom percentage. As illustrated in Figure S1 in the Supporting Information, after calcination treatment for the doped samples there is a notable color change, which ranges from sky-blue to various other colors; this is distinct from the white undoped TiO2 powder. The doped TiO2 powder samples display various colors that depend on the type of metal ions; Fe3+, Co2+, and Ni2+-doped samples are chalky yellow, clay bank, and yellowish-green, respectively.

It is desirable to maintain the integrity of the crystal structure of the photocatalytic host material to produce favorable changes in the electronic structure. The crystallographic structures of the resultant materials are confirmed by X-ray diffraction (XRD). All of the diffraction peaks match well with the crystal structure of the anatase-phase TiO2 (space group: I41/amd, Joint Committee on Powder Diffraction Standards No. 21-1272) (Figure 1 A). No crystalline phase for the other metal oxides is observed, even if the M/Ti ratio is as high as 2 % (Figure S2, Supporting Information). The TEM images of the M-0.75 samples are shown in Figure 2. All of the samples display a lateral size of 90–100 nm and a thickness of 15–20 nm (Figure 2 A–D). Well-defined sheet-shaped TiO2 structures are clearly observable in the higher magnification image of Figure 2 E. A high-resolution TEM image (Figure 2 F) indicates part of a single nanosheet with visible lattice fringes. Two set of lattices are presented. These are oriented perpendicular to each other with an equal lattice spacing of 0.19 nm, which corresponds to the (2 0 0) and (0 2 0) atomic planes.24 The same region of the fast Fourier transform (FFT) pattern is indexed into the diffraction spots of the [0 0 1] zone, as designated in the Figure 2 F inset. The clear diffraction spots indicate the single crystalline nature of the nanosheets. Based on the above structural information, the percentage of the highly active {0 0 1} facets in the samples are estimated to be approximately 74 % (refer to our previous report25 for detailed calculations). Surprisingly, the size of M-0.75 is almost identical regardless of the different types of dopants used, which indicates that the type of doping ion has little influence on the growth of TiO2 nanosheets. There is no significant difference in the size of anatase nanosheets compared to that of pure TiO2 if the dopant concentration is decreased to 0.5 or 0.25 % (Figure S3 A and B, Supporting Information). In contrast, if the dopant concentration is increased to 1 %, the lateral size of samples decreases to approximately 83 nm, but the thickness increases to approximately 20 nm (Figure S3 C). Also, the percentage of reactive facets is accordingly decreased to 62 %. If the dopant concentration is further increased to 2 %, the side length and the thickness of the TiO2 nanocrystals are found to be approximately 45 and 25 nm, respectively (Figure S3 D). Unfortunately, the percentage of {0 0 1} facets is further decreased to 36 %. The size (b), thickness (h) and percentage of {0 0 1} facets in all the samples are listed in Table S1 in the Supporting Information. These data indicate the adverse role of the dopant concentration in the isotropic growth of nanosized TiO2 single crystals with exposed {0 0 1} facets. Moreover, many tiny pores in all the TiO2 nanosheets are found, which can be attributed to the hydrofluoric acid etching effect.26

thumbnail image

Figure 1. A) XRD patterns of the M-0.75 and pure TiO2 samples. B) Photographs of M-0.75 and pure TiO2.

Download figure to PowerPoint

thumbnail image

Figure 2. TEM images of A) pure TiO2, B) Fe-0.75, C) Co-0.75, D) Ni-0.75, and E) an individual nanosheet. A high-resolution TEM image of F) Ni-0.75 is shown in the inset of E. Inset: FFT pattern.

Download figure to PowerPoint

As for the dependence of the formation mechanism on the dopant concentration, we assume that the introduced nitrate ions (NO3) act as an absorbing agent for the synthesis of TiO2 nanosheets. After a Ti precursor and a trace amount of metal nitrates is added into the isopropyl alcohol solution, the nitrate ions adsorb onto the surface of the nucleus and reduce the absorbing capacity of the fluoride ions, which can markedly reduce the surface energy of the {0 0 1} facets, thus accelerating the anisotropic growth of TiO2 crystals and decreasing the percentage of {0 0 1} facets. The more nitrate ions that are added, the more apparent the phenomenon of the changed percentage of {0 0 1} facets in TiO2 nanosheets that are observed. By contrast, if no metal nitrate is introduced, the higher absorbing capacity of fluoride ions on the surface makes the isotropic growth more apparent; a similar phenomenon was reported by Ma et al.27 In addition, as summarized in Table S1, 0.75 % is a critical point.

It had been demonstrated that the metal ion and doping levels can greatly influence the property of the host materials.7 Thus, we proceeded to examine the chemical composition and doping level of the colored powder by using X-ray photoelectron spectroscopy (XPS) analysis (Figure S4, Supporting Information). As no peak at a binding energy of 684 eV, which corresponds to the F 1s orbital, is observed owing to calcinations, this rules out the existence of F ions. The atom ratios of M to Ti in the resulting samples are summarized in Table S1. We chose M-0.75 to analyze in detail because the high-resolution XPS spectra of same-element ion doped TiO2 are similar (Figure 3). The Fe 2p, Co 2p, and Ni 2p high-resolution XPS spectra of M-0.75 are shown in Figure 3 A–C , respectively. The spectrum in the Fe 2p3/2 region of the corresponding sample indicates that the peak at 710.6 eV is symmetrical and can be ascribed to Fe3+ ions. This is not difficult to understand because the dopant is Fe(NO3)3. Thus, the Fe element in the samples exists mainly in the +3 oxidation state (Fe3+).28 The Co 2p core level spectrum of the Co-doped sample is shown in Figure 3 B. Four broad peaks in the Co 2p region can be observed, which are the Co 2p3/2 and Co 2p1/2 peaks located at 780.2 and 796.6 eV, respectively. Also, two satellite peaks are centered at 786.3 and 802.1 eV, which can be attributed to Co2+ ions.9 The Co metal phase in the doped samples is completely ruled out, because the metallic Co 2p3/2 energy level has a peak position at 778.3 eV.29 The satellite peaks (denoted as “Sat.”) on the high-energy side of the principal Co 2p1/2 and Co 2p3/2 lines are typically characteristic of high-spin Co2+ species.30 The spectrum is characterized by an Ni 2p3/2 peak at 855.4 eV and an Ni 2p1/2 peak at 873.0 eV. The difference between these peaks, namely, 17.6 eV, suggests the existence of a Ni[BOND]O band.10 No presence of other oxidation states or a metallic state of Ni is found. The appearance of satellite structures implies the presence of a high-spin divalent Ni2+ state in the sample.31 Moreover, the oxidation state of Ti in the undoped sample (Figure 3 D) shows two bands in this region, which are located at 464.0 and 458.2 eV. These are attributed to the Ti 2p1/2 and Ti 2p3/2 energy levels, respectively. The splitting difference of 5.8 eV between the two bands indicates the existence of the Ti4+ species in the sample.32 The two peaks of the Ti 2p energy levels in M-0.75 (Figure S5 A) are slightly shifted toward higher energies compared to the values of undoped TiO2. This implies that the metal ions may be incorporated into the TiO2 lattice.33 Furthermore, the order of the shifting value is Ni>Co>Fe. Owing to the difference in electronegativity (Fe=1.83, Co=1.88, Ni=1.91 vs. Ti=1.54), a trace quantity of Ti ions may transform to a higher valent state by releasing excessive electrons and thereby achieving charge balance in the TiO2 lattice, following the incorporation of metal ions. This is possibly the reason that the XPS peaks in doped TiO2 shift toward higher energies. The binding energy of the O 1s peak of TiO2 (Figure S5 B) is approximately 529.63 eV, and this is ascribed to the O[BOND]Ti4+ interaction. As seen by comparing the doped samples, the above peaks shift to lower energy regions, and the order of the shifting value is Ni>Co>Fe. This can be attributed to the different variations in electronegativities between the Ti- and Fe-group elements, and it confirms the formation of a M[BOND]O[BOND]Ti complex in doped powders.34 The shoulder centered at 531.7 eV in all the samples is attributed to the OH groups that are absorbed onto the surface,35 which can react with the photoinduced holes and yield surface-bound OH radicals with high oxidation reactivity. The results of the XPS spectra from the surface of the samples reveals that the Fe-group ions are incorporated into the anatase TiO2 lattice, and Ti4+ ions are substituted with foreign ions.

thumbnail image

Figure 3. High-resolution XPS spectra of the Fe-0.75, Co-0.75, Ni-0.75, and pure TiO2 samples for A) Fe 2p, B) Co 2p, C) Ni 2p, and D) Ti 2p peaks.

Download figure to PowerPoint

Photocatalytic degradation of dyes with doped TiO2

Dyes are important organic compounds that are widely applied, for example, in the textile or food industries, in which they are often regarded as environmental contaminants. To address this, the present work evaluates the photocatalytic activity of synthetic samples by monitoring the degradation of MB under UV light (λ<400 nm) and visible light irradiation (λ>420 nm). Although all doped TiO2 samples show photocatalytic activity in the presence of UV light, a large difference is observed between the doped TiO2 samples, which depends on the doping level (Figure S6 A, C, E). The activities of M-0.25, 0.5, and 0.75 are far greater than those of M-1 and 2, and the same trends occur in the three types of doped photocatalysts. To compare the photocatalytic activity of the samples directly, we summarize the degradation efficiency after 5 min under UV light illumination in Figure 4 C and Table S1. We find that the photocatalytic activities increase with the dopant concentration if the M/Ti ratio is below 0.75 %. However, the photocatalytic activities of the samples decrease and are even slightly lower than that of undoped TiO2 as the M/Ti ratio increases further from 1 to 2 %; this indicates that M-0.75 is the optimal doped amount in our system. More interestingly, the photocatalytic activities of Ni-doped TiO2 are much higher than Co- and Fe-doped TiO2, and the order of activity is Ni->Co->Fe-doped TiO2. As far as we know, this is the first time that the order of photocatalytic activities of Fe-group element ion-doped TiO2 nanosheets is shown.

thumbnail image

Figure 4. Photocatalytic degradation of MB in the presence of M-0.75 and pure TiO2 nanosheets under A) UV and B) visible-light irradiation, respectively. Variation of the degradation efficiency of M-doped TiO2 at different atom ratios and pure TiO2 nanosheets after C) UV and D) visible-light irradiation, for 5 and 180 min, respectively.

Download figure to PowerPoint

To confirm this conclusion, we examined the degradation of MB under visible light (Figure S6 B, D, F). If no catalyst is added, the degradation of MB is approximately 30 %. However, approximately 90, 94, and 96 % of MB is eliminated from the solution in 3 h in the presence of Fe-0.75, Co-0.75, and Ni-0.75, respectively. The degradation efficiency of other doped TiO2 samples after 3 h in the presence of visible light is summarized in Figure 4 D. The same tendency is observed as with the UV light illumination. For comparison, the undoped TiO2 was examined and it is found that approximately 85 % of MB decomposes under the same testing conditions. It is known that anatase TiO2 is a wide-band-gap semiconductor (3.2 eV) that absorbs an excitation wavelength below 387 nm. The observed photocatalytic activity of undoped TiO2 nanosheets under visible light irradiation should be attributed to the photosensitization degradation of MB, which extends the absorption of TiO2 to the visible light region.36 The self-sensitized degradation mechanism of MB on pure TiO2 in the presence of visible light is summarized in Figure S7 in the Supporting Information. The reaction mechanism of the whole degradation process on anatase TiO2 is similar to the electron transfer mechanism in dye-sensitized TiO2 solar cells. If MB can be excited under visible light, the excited electrons are first injected into the conduction band of TiO2, and then are captured by O2 to generate O2 ions. Finally, MB is decomposed by the active oxygen species.

Insights into the photocatalytic activity of doped TiO2

As far as we know, the photocatalytic activity of a semiconductor is largely controlled by: 1) the light absorption properties, such as the light absorption spectra and coefficients; 2) the charge transportation and separation rate; and 3) the available specific surface area.7 Firstly, in our case, the specific surface areas of all samples are almost identical, as summarized in Table S1. Then, in the presence of UV light, there is a negligible difference between the light absorption properties of these TiO2 nanosheets owing to the light absorption of all samples over the whole UV region. Thus, the charge transportation and separation rate is the critical factor in the photocatalytic properties of the samples. To unravel the details of the photocatalytic activity that depend on the charge transportation and recombination rate, we conducted electrochemical impedance spectra (EIS), photocurrent, and photoluminescence (PL) tests.

In previous studies, an EIS analysis was commonly used to clarify the above results.3738 The EIS Nyquist plots of undoped TiO2 and the M-0.75 sample with UV irradiation are shown in Figure 5 A. The arc radius on the EIS Nyquist plot of Ni-0.75 is the smallest of the four samples, and the order of the arc radius is Ni-0.75<Co-0.75<Fe-0.75<TiO2. In the EIS Nyquist plot, the arc radius on the EIS spectra indicates the reaction rate that occurs at the surface of the material; the smaller semicircle size reflects an effective separation of photogenerated electron–hole pairs and a fast interfacial charge transfer to the electron donor or acceptor.39 Thus, more effective separation of photogenerated electron–hole pairs and faster interfacial charge transfer occurs on the Ni-0.75 photocatalyst under this condition. Compared with the other photocatalysts (Figure S8, Supporting Information), it is found that if the atom percentage of dopant is equal, the arc radius of Ni-doped TiO2 is the smallest, and that of the Co-doped sample is smaller than the Fe-doped sample and pure TiO2. On the other hand, the semicircle size increases with the dopant concentration if the original atom ratio of M and Ti is below or above 0.75 %, if a constant type of dopant is employed. A special case occurs if the original molar ratio of M to Ti reaches 2 %; the arc radius of M-2 is even higher than that of pure TiO2 nanosheets. These results indicate that the Ni-0.75 sample is able to enhance the separation of photogenerated electron–hole pairs.

thumbnail image

Figure 5. A) EIS spectra, B) photocurrent curves, C) PL emission spectra, and D) diffuse reflectance spectra of M-0.25 and TiO2 nanosheets.

Download figure to PowerPoint

To further support the above proposition, transient photocurrent responses were recorded for photoelectrodes that consisted of all samples under UV irradiation. The I–t curves for the M-0.75 sample and TiO2 with UV light irradiation are shown in Figure 5 B. It is implied that the photocurrent is determined largely by the separation efficiency of the photogenerated electron–hole pairs within the photocatalyst. The photocurrent of the M-doped TiO2 is higher than that of pure TiO2, which indicates that the separation efficiency of photoinduced electrons and holes can be improved through the electronic interaction between the doping metal ion and TiO2 nanosheets. In addition, the photocurrent of Ni-doped TiO2 is the highest in the test samples, and that of the Co-doped sample is higher than that of the Fe-doped TiO2 powder. The photocurrent of the rest of the samples is shown in Figure S9 in the Supporting Information. The trend is the same as with the M-0.75 samples, and the photocurrent of M-doped TiO2 first increases with the M/Ti ratio; it reaches a maximum value at M/Ti=0.75 %, and then decreases. This trend further confirms the results from EIS testing.

PL emission spectra may be useful for understanding the behavior of photoexcited electrons and holes in our samples, because PL emissions result from the recombination of free carriers.40 PL signals for the M-0.75 sample and TiO2 under excitation at λ=325 nm are provided in Figure 5 C. The lowest PL intensity for Ni-0.75 indicates the lowest recombination rate of electrons and holes, which is consistent with the photocatalytic activity of the sample. In addition, the remainder of the results also indicate that the PL intensity of Ni-doped TiO2 is the lowest except for the Ni-2 sample (Figure S10, Supporting Information), which is slightly higher than that for undoped TiO2. More importantly, the PL results demonstrate that the order of charge transportation and separation in doped TiO2 is Ni->Co->Fe-doped TiO2, and the 0.75 % atom percentage of the dopant has the best result; this is in good agreement with the results of EIS and the photocurrent tests. All of the above results indicate that the photocatalytic properties of the samples are directly determined by the charge transportation and separation rate.

What is the effect of the type of doped ion on the charge transportation and separation rate? According to a study by Shah et al.,41 the position of the dopant can be determined by the size differences between the ionic radius of the Ti4+ ion and the dopant’s ionic radius. If the ionic radius of the dopant is larger than that for Ti4+, it would be likely to enter the lattice by interstitial means, which would then result in a large mismatch of the lattice. To compensate for the mismatch, the creation of an oxygen vacancy in the surface of the TiO2 lattice would be likely, which would enhance the electron trapping efficiency.42 Owing to the larger difference in ionic radius between Co and Ti compared with the difference with Fe or Ni (Fe3+=64.5 pm, Co2+=74.5 pm, Ni2+=69 pm vs. Ti4+=60.5 pm), then in theory, more oxygen vacancies in Co-doped TiO2 should be formed than with Fe and Ni doping. However, the highest charge transportation and separation rate is observed in Ni doping.

This phenomenon might also be related to the electronic structure of the doped-TiO2. If TiO2 is doped with Fe or Co, the electrons are localized around each dopant, but the electrons from the Ni dopant are delocalized.20 As previously reported, the photoreactivity of ion-doped TiO2 appears to be a complex function; it is influenced by various factors such as the energy level of dopants within the TiO2 lattice, their d electronic configurations, or the electron donor concentrations.16, 43 Thus, the Ni doping ion is more appropriate than the Fe or Co ion in our case. How, then, can we understand the relationship between the recombination rate and the dopant concentration? In a previous report, Choi et al.16 found that if the dopant concentration is below the optimal value, the photoreactivity increases with an increasing dopant concentration, because there are fewer trapping sites available to suppress the recombination of electrons and holes. However, for a higher dopant concentration beyond the optimal value, the excessive trapping sites become efficient recombination centers because the average distance between the trap sites decreases with an increasing number of dopants confined within a particle. This increases the recombination rate.44 Therefore, the relative recombination rate of a metal ion dopant depends on whether it serves as an effective trap or as a recombination center, and the dual role of the doping ion can be converted depending on the dopant concentration. If a value is reached above the degradation efficiency of pure TiO2 nanosheets, the positive effective trap plays a dominant role if the atom ratio of M/Ti is below 1 % (Figure 4 C). In contrast, if the M/Ti ratio is increased continually, the negative recombination center plays a major role. In addition, the decreasing {0 0 1} active facets, owing to an increased dopant concentration, is an important factor in the declining photoactivity. Based on the above results, in our case Ni-0.75 has the optimal effective trap and photocatalytic activity under UV irradiation.

If the photocatalytic decomposition testing is operated under visible light irradiation, the light absorption ranges of the catalysts play a necessary role. The UV/Vis diffuse reflectance spectra of the pure TiO2 nanosheets and M-0.75 powders are shown in Figure 5 D. The presence of metal ions significantly affects the optical property of light absorption for M-doped TiO2. The addition of metal ions induces the shifting absorption edge to the lower energy region, as observed in all of the doped TiO2 samples with different addition ratios of metal ions (Figure S11, Supporting Information). It is well-known that the origin of the visible spectra in the metal-doped sample owes to the formation of a dopant energy level within the band gap of TiO2.10 The redshift of the band edge absorption threshold benefits by the electronic transitions from the valence band to the dopant level or from the dopant level to the conduction band.45 Therefore, the observed redshift in the band gap implies that the optical absorption of the TiO2 nanocrystals can be tuned from the UV to the visible light region through metal ion doping. Notably, the band gap of the Ni-doped sample is the narrowest, and that of the Fe-doped TiO2 sample is larger than that of the Co-doped TiO2 sample, which might be attributed to the different position of the three dopant energy levels within the band gap of TiO2.20 The additional absorptions observed in the visible region may owe to a sub-band-gap transition that corresponds to the excitation of 3d electrons in the Fe-group ion to the TiO2 conduction band, and the d–d transition of the metal ion in the TiO2 matrix.9, 10, 28 The approximately estimated band gaps are summarized in Table S1. Interestingly, a decreased band gap seems to depend on an increase in the doping ion content. The reason might be that the doping metal ions can form a dopant level near the valence band of TiO2, and thus reduce the band gap of TiO2.46

To confirm the efficiency of the visible light response, the transient photocurrent measurements under visible light irradiation were performed. It is no surprise that little photocurrent can be detected in the presence of visible light, which indicates that the photogenerated electron–hole pairs cannot be effectively separated under the excitation wavelength of λ>420 nm. That is, this range of light contributes little to the catalytic degradation of MB. Therefore, the degradation of MB under visible light can be attributed to the dye-photosensitization effect. What kind of role do the metal ions play in this system? Further analysis of M-doped samples is consistent with our prediction. The redox potentials E0(Fe3+/Fe2+)=+0.77 V, E0(Co2+/Co)=−0.28 V, and E0(Ni2+/Ni)=−0.25 V lie between the conduction-band potential of TiO2 (Ecb=−0.5 V vs. the normal hydrogen electrode) and the valence-band potential (Evb=2.7 V vs. the normal hydrogen electrode). In this situation, the Mn+/M(n–1)+ species trap the electrons in the conduction band of TiO2, which migrate from the LUMO energy band of the MB and thus contribute to the separation of the electron–hole pairs.47 The detailed explanation is as follows. The adsorbed MB is excited under the illumination of visible light (Figure 6). Then, the electrons migrate from the dye to the conduction band of the TiO2 matrix, and subsequently transfer easily to the surface energy level of the M species. The electron acceptors such as O2 that are adsorbed onto the catalyst surface or dissolved in water can immediately react with these electrons to generate O2.− ions. Meanwhile, the cationic dye radicals MB.+ also react with hydroxyl ions to generate a hydroxyl radical HO.−; the resulting strong oxidants O2.− and HO.− can oxidize MB to yield the products. In the results of the EIS and photocurrent tests, Ni-doped TiO2 prepared by the current approach has the highest charge transportation and separation rates. Therefore, Ni-doped TiO2 shows the highest visible-light photocatalytic degradation activity, and Co- or Fe-doped TiO2 shows comparable visible-light photodegradation rates. In addition, the effect of the dopant concentration in M-doped TiO2 with visible irradiation is the same as that of M-doped TiO2 under the illumination of UV light.

thumbnail image

Figure 6. Representation of the photocatalytic processes on M-doped TiO2 surfaces under visible-light irradiation.

Download figure to PowerPoint

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and Discussion
  5. Conclusions
  6. Experimental Section
  7. Acknowledgements
  8. Supporting Information

In summary, we prepared a series of Fe-group ion-doped anatase TiO2 with exposed {0 0 1} facets, for which the dopant concentration was varied from 0.25 to 2 %. The size and percentage of the {0 0 1} facets in doped TiO2 nanosheets is almost identical if the dopant concentration is increased to 0.75 %, but the size drastically decreases with a further increase in the doping ion content. The synthesized colored powder shows different photocatalytic activities that depend on the type and concentration of additional doping ions, which can be attributed to a significant impact on the charge transportation and separation rate. In contrast with Fe- and Co-doped TiO2, the Ni-doped TiO2 contains a higher charge transportation and separation rate, which leads to higher photocatalytic activity under UV or visible-light irradiation. Moreover, the dual role of the doping ions is shown, and it can be converted because this depends on the doping concentration. Importantly, this is the dominant factor in the photocatalytic activity of the doped TiO2 photocatalyst. These studies are expected to further cement our understanding of the effects of similar element doping on the photocatalytic property of a TiO2 host, and enable us to design efficient ion-doped TiO2 photocatalysts for environmental remediation and energy conversion.

Experimental Section

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and Discussion
  5. Conclusions
  6. Experimental Section
  7. Acknowledgements
  8. Supporting Information

Preparation of undoped and Fe-, Co-, and Ni-doped TiO2 nanosheets

An improved fluoride-mediated self-growing method was used to produce pure, Fe-, Co-, and Ni-doped TiO2 nanoshells.48 In a typical process, appropriate amounts of Fe(NO3)3, Co(NO3)2, and Ni(NO3)2 were added to isopropyl alcohol (30 mL), respectively. Then, the mixture was ultrasonicated for 10 min (500 W power) and tetrabutyl titanate (2 mL) was added dropwise into the above solution and ultrasonicated for another 10 min. The mixed solution and HF (0.35 mL, 40 %) were transferred to a polytetrafluoroethylene (Teflon) lined stainless steel autoclave (50 mL), and then placed in an electric oven at 200 °C for 24 h. The autoclave was then removed from the oven and left to cool naturally to room temperature. The precipitate was collected by centrifugation at 12 000 rpm for 20 min, washed thoroughly with ethanol several times, and fully dried at 40 °C in an oven to obtain a sky-blue powder. All of the products were calcinated at 600 °C for 3 h with a heating rate of 5 °C min−1 to obtain a colored and highly crystalline anatase phase.

Photoreactivity experiments

In a typical experiment, the photocatalysts (30 mg, undoped and doped TiO2 nanosheets) were suspended in MB (30 mL, 10 ppm) solution. Before irradiation, the suspension was ultrasonicated for 5 min and stirred for 1 h in the dark to achieve adsorption–desorption equilibrium. Under ambient conditions and stirring, the solution was irradiated by using a 500 W power Xenon lamp (LSXS-500) with a visible cutoff filter (λ<400 nm) and a UV cutoff filter (λ>420 nm) (visible-light test), respectively. A sample solution (3 mL) was taken during a certain time interval during the experiment and centrifuged to remove the catalyst completely. The solution was analyzed on a UV/Vis spectrophotometer (PerkinElmer Lambda 25). For comparison, blank experiments without a catalyst were also performed.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and Discussion
  5. Conclusions
  6. Experimental Section
  7. Acknowledgements
  8. Supporting Information

Financially supported by the National Natural Science Foundation of China (No. 2177142) and the Important Deployment Project of Chinese Academy of Sciences (KZZD-EW-TZ-16).

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and Discussion
  5. Conclusions
  6. Experimental Section
  7. Acknowledgements
  8. Supporting Information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
cctc_201300879_sm_miscellaneous_information.pdf1183Kmiscellaneous_information

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.