SEARCH

SEARCH BY CITATION

Keywords:

  • copper;
  • hydrogen;
  • mesoporous materials;
  • titanium;
  • water splitting

Abstract

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

Cu2O-decorated mesoporous TiO2 beads (MTBs) are developed as a low-cost, highly efficient photocatalyst for H2 production. MTBs with a high specific surface area of 189 m2 g−1, a large pore volume of 0.43 cm3 g−1 and a suitable pore size of 8.9 nm are decorated with band-structure-matched Cu2O nanocrystals through a simple, fast and low-cost chemical bath deposition process. The Cu2O nanocrystals serve as an electron–hole separation centre to promote H2 evolution. Under optimal operation conditions, an ultra-high specific H2 evolution rate of 223 mmol h−1 g−1 is achieved. The success is attributed to the structural advantages of the MTBs of high specific surface areas, large pore volumes and suitable pore sizes together with the much improved electron–hole separation and light utilisation of the Cu2O-decorated MTBs. The H2 evolution rates achieved with the Cu2O-decorated MTBs are one order of magnitude higher than those achieved by commercial P25 TiO2.


Introduction

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

Global warming is a serious issue that has drawn a great deal of research attention in recent years.1, 2 The main fuel or energy sources used today are carbon-containing materials such as coal, petroleum and natural gases that generate CO2 on combustion, which leads to severe greenhouse effects.3, 4 To reduce the evolution of greenhouse gases, alternative, clean energies are under investigation, which include piezo-electric,5 solar6, 7 and wind energies.8 H2, the cleanest energy, produces only water on combustion. H2 production has thus become a research area of intensive and extensive investigation.911 Particularly, H2 evolution through photocatalytic water splitting is an attractive way to produce H2 as it uses mainly sunlight and water, both renewable and widely available.1214

In photocatalytic water splitting, besides the light source, the evolution of H2 is significantly affected by the properties of the photocatalyst. There are two basic requirements for photocatalysts applied in water splitting. First, a band gap larger than the energy difference of the reduction and oxidation potentials of water (≈1.23 eV) accompanied by a suitable band structure is required. The conduction band of the semi-conductor photocatalyst must be more negative than the reduction potential of water, and the valence band must be more positive than the oxidation potential of water to enable water decomposition for the generation of H2 and O2.15 Charge separation efficiency is another important issue for the performance of photocatalytic water splitting.16, 17 Upon illumination with light that possesses high enough energy to overcome the bandgap of the photocatalyst, the photocatalyst would be optically excited to produce electron–hole pairs. However, the photoinduced electrons may not be able to migrate to the surface of the materials to proceed with the necessary reduction for H2 production.18 Instead, the electrons may re-combine with holes and thus are removed from the system. Therefore, semi-conductor materials with a suitable band structure and excellent electron–hole separation are sought for application in photocatalytic water splitting.

TiO2 is a non-toxic, low-cost, photochemically stable, Earth-abundant semi-conductor material with a band gap of 3.2 eV that has been extensively applied in photocatalysis.19, 20 The band structure of anatase TiO2 meets the requirements for photocatalytic water splitting and anatase TiO2 has thus attracted a great deal of research attention in H2 evolution. For photocatalytic water splitting, the H2 evolution reaction occurs at the interface between the photocatalyst and reactant solution. Therefore, it is important to increase the surface area of the TiO2 material to accommodate more active sites for the reactions involved and create adequate porous structures for the fast mass transfer necessary for the reactions.21

As a result of the poor electron mobility of TiO2, the photoinduced electrons are prone to electron–hole re-combination. Therefore, there is still ample room for improvement in H2 production efficiency if TiO2 is used as the photocatalyst.22 To improve the performance of TiO2, some researchers tried to load noble metals as a co-catalyst onto the surface of TiO2. The noble metal can attract electrons to improve the electron–hole separation, which is beneficial for H2 production. Pt is probably the most studied and most often used metallic co-catalyst.23, 24 Furthermore, some researchers developed hetero-junctions in the photocatalyst to promote electron–hole separation through suitably matched band structures.2527 Cu2O is a semi-conductor with a small band gap of about 2 eV. It has been applied in photocatalysis28, 29 and coupled with oxide semi-conductors with larger band gaps for the degradation of pollutants. Nevertheless, the number of studies on Cu2O/TiO2 composites for applications in photocatalytic water splitting for H2 production is still quite limited and there is still ample room for performance improvement, possibly because of the inadequate structural design of TiO2.3033 It has to be stressed that, for the same system of materials, a different structural design can greatly affect the functional performance of the system.

In this study, we develop a simple metal-salt-based hydro-thermal process to prepare sub-micron-sized mesoporous TiO2 beads (MTBs) with high specific surface areas, large pore volumes and suitable pore sizes for applications in photocatalytic water splitting. The structural features of the MTBs are particularly suited for applications that involve heterogeneous reactions/interactions such as in the photoanodes of dye-sensitized solar cells34 and photocatalysts for pollutant degradation35 and H2 evolution. The high specific surface area provides a large number of active sites to host the heterogeneous reaction/interaction events, whereas the large pore volume and suitable pore size facilitate the fast mass transfer of reactants and products, critical for the heterogeneous reactions/interactions. In addition, the sub-micron size of the MTBs enables easy recycling of the photocatalyst. The MTBs serve as an excellent base material for the decoration of Cu2O nanocrystals, which is achieved with a simple chemical bath deposition process. The chemical bath deposition process is able to uniformly deposit Cu2O onto the surfaces of the constituent nanocrystals of the MTBs to give a high extent of the utilisation of the deposited Cu2O to affect the charge separation, essential for highly efficient H2 evolution. There is an optimal Cu2O loading with respect to H2 evolution rates. The present low-cost process creates a photocatalyst that exhibits excellent charge separation and thus ultra-high H2 evolution rates. Under optimal operation conditions, an ultra-high specific H2 evolution rate of 223 mmol h−1 g−1 is achieved. This new approach to combine the advantageous structural feature of a base semi-conductor with the much improved charge separation, made possible through the decoration of a second band-structure-matched semi-conductor, proves its merits in the development of semi-conductor photocatalysts for water splitting.

Results and Discussion

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

The MTBs are formed through aggregation of the constituent TiO2 nanocrystals accompanied by inter-nanocrystal growth and through etching of the MTBs with Cl.34 These MTBs are mostly granular in shape with an average size of around 840 nm (Figure 1 a). From the high-magnification SEM image shown inset in Figure 1 a, it can be seen that the MTBs are composed of nanocrystals of around 10 nm in size. Once the Cu2O is deposited onto the MTBs, the boundaries between the constituent nanocrystals of the MTBs become blurred, which indicates the occupancy of the inter-nanocrystal space with Cu2O deposits. On increasing the concentration of the Cu2O precursor, the amount of Cu2O deposits increases, which gradually forms a Cu2O coating layer on the outer surfaces of the MTBs and eventually covers the TiO2 nanocrystals underneath (inset in Figure 1 c and d). At a high precursor concentration of 0.3 M, the Cu2O deposits can even connect neighbouring MTBs to form large MTB clusters (Figure 1 d). The constituent nanocrystals of the MTBs are strongly connected together and cannot be broken apart even under strong ultra-sonication. These as-prepared MTBs have good crystallinity of the anatase phase as can be seen from the well-resolved diffraction peaks and the excellent match of the diffraction pattern with the database pattern of JCPDS # 89-4921. Based on the (1 0 1) diffraction peak, the grain size of the MTB is estimated to be 8.4 nm from the Scherrer equation. The grain size is roughly equal to the nanocrystal size, which implies that the constituent nanocrystals are in fact single crystalline, which is beneficial for fast charge transport within the MTB.

thumbnail image

Figure 1. SEM images of a) MTBs and Cu2O/MTB composite photocatalysts loaded with b) 0.1, c) 0.2 and d) 0.3 M Cu2O precursors.

Download figure to PowerPoint

Upon deposition of Cu2O, several extra diffraction peaks, which include (1 1 1), (2 0 0) and (2 2 0) of Cu2O (JCPDS # 65-3288), appear in the XRD patterns of the Cu2O-decorated MTBs (Figure 2). These extra diffraction peaks become more pronounced with the increasing concentration of the Cu2O precursor. The Cu2O loading in the MTBs was analysed by energy dispersive spectroscopy (EDS) measurements. The measurement covers an area of micro-meter scale to acquire representative data. The results are summarised in Table 1 in terms of the atomic percentage of Cu. The mole percentage of Cu2O increases from 1.15 to 1.95 then to 2.60 with increasing concentration of the Cu2O precursor from 0.1 to 0.2 and 0.3 M.

thumbnail image

Figure 2. XRD patterns of bare MTBs and Cu2O/MTB composites prepared at three different Cu2O precursor concentrations.

Download figure to PowerPoint

Table 1. Cu concentration of Cu2O/MTB composites prepared at three different Cu2O precursor concentrations.
SampleCu [mol %]
MTB
Cu2O/MTB-0.1 M1.15
Cu2O/MTB-0.2 M1.95
Cu2O/MTB-0.3 M2.60

To further confirm the formation of Cu2O in the decorated MTBs, X-ray photoelectron spectroscopy (XPS) was conducted, and the results are shown in Figure 3. Two characteristic binding energies at 952.7 and 932.5 eV can be identified for Cu 2p1/2 and Cu 2p3/2, respectively. However, this set of binding energies can be attributed to either Cu0 or CuI as the binding energies of the two states differ only by 0.1 eV,36 which is within the experimental error. Nevertheless, the possibility of CuII can be safely excluded as the characteristic binding energies of CuII for Cu 2p1/2 and Cu 2p3/2 are 962 and 942.2 eV, respectively.37 Together with the XRD analyses, we can confirm that Cu2O is successfully introduced onto the surfaces of the constituent nanocrystals of the MTBs with the chemical bath deposition process.

thumbnail image

Figure 3. XPS spectra of Cu2O/MTB composites prepared at three different Cu2O precursor concentrations.

Download figure to PowerPoint

The micro-structural characteristics of the MTBs, plain or Cu2O-decorated, were further investigated with N2 adsorption–desorption analysis, and the results are shown in Figure 4. All four samples exhibit a type IV isotherm with an evident type H3 hysteresis, typical for mesoporous materials. This observation also indicates that the decoration of Cu2O onto the MTBs does not change the essential micro-structure of the MTBs. The corresponding structural parameters, which include specific surface areas, average pore sizes and pore volumes, are summarized in Table 2 for comparison. The bare MTBs possess a high specific surface area of 189 m2 g−1, a large pore volume of 0.43 cm3 g−1 and a suitable average pore size of 8.9 nm. The high specific surface area can accommodate a large amount of active sites for the photocatalytic reactions, whereas the large pore volume and suitable pore size allow fast mass transfer within the MTBs for the involved reactants and products. Upon Cu2O decoration, the specific surface areas and pore volumes of the MTBs decrease with increasing concentration of the Cu2O precursor, whereas the average pore sizes remain roughly unchanged at around 9 nm (Table 2).

thumbnail image

Figure 4. N2 adsorption–desorption isotherms of commercial P25 TiO2 and Cu2O/MTB composites prepared at three different Cu2O precursor concentrations.

Download figure to PowerPoint

Table 2. Micro-structural parameters of P25 TiO2, MTBs and Cu2O/MTB composites prepared at three different Cu2O precursor concentrations.
SampleSpecific surfaceAverage porePore volume
area [m2 g−1]size [nm][cm3 g−1]
P255527.40.375
MTB1898.90.430
Cu2O/MTB-0.1 M1828.60.396
Cu2O/MTB-0.2 M1679.10.382
Cu2O/MTB-0.3 M1359.60.347

The UV/Vis absorption spectra of the bare and Cu2O-decorated MTBs are shown in Figure 5 along with the spectra of commercial P25 TiO2 and Cu2O powders for comparison. The P25 TiO2 and bare MTBs show onset absorption at wavelengths shorter than 400 nm in the UV region, as expected, whereas the Cu2O powders exhibit onset absorption at around 642 nm, which corresponds to a band gap of 1.93 eV. For the Cu2O-decorated MTBs, the extended adsorptions in the visible-light region clearly come from the contribution of the Cu2O nanocrystals. Evidently, a higher extent of visible-light absorption is achieved for the Cu2O-decorated MTBs with the increasing concentration of the Cu2O precursor.

thumbnail image

Figure 5. UV/Vis spectra of commercial P25 TiO2, Cu2O, MTBs and Cu2O/MTB composites prepared at three different Cu2O precursor concentrations.

Download figure to PowerPoint

The amount of H2 evolution from photocatalytic water splitting is ideally a linear function of the amount of photocatalyst contained in the system. More photocatalyst provides more active sites to proceed with the H2 evolution and thus larger amounts of H2 production. Nevertheless, too much photocatalyst may cause light shielding that reduces the extent of utilisation of the photocatalyst and thus the amount of H2 evolution. To check the effects of the photocatalyst concentration on light shielding, different amounts of photocatalyst, 0.0005, 0.01 and 0.05 g, were added into 500 mL deionised (DI) water, and their transmittance spectra were investigated. The results are shown in Figure 6. Two sets of samples were measured and compared, one P25 TiO2 and the other Cu2O/MTB composite. Here, P25 TiO2 is taken as the comparison base as it is a well-known photocatalyst that exhibits excellent performance in both photocatalytic pollutant degradation and photocatalytic water splitting. First, evidently, for both sets of samples, the transmittances are significantly reduced on increasing the amount of the photocatalyst from 0.0005 to 0.01 then to 0.05 g. Also presented in Figure 6 are photos of the corresponding suspensions with plain DI water (far right) included for comparison. It is evident from the photos that the cloudiness of the suspensions increases significantly with increasing photocatalyst concentration, in agreement with the UV/Vis spectra. These suspensions are stable and remain well dispersed even left undisturbed for hours.

thumbnail image

Figure 6. a) Transmittance spectra and b) photos of P25 TiO2 (top set) and Cu2O/MTBs (bottom set; prepared at a Cu2O precursor concentration of 0.2 M) suspensions with different photocatalyst concentrations.

Download figure to PowerPoint

For the most dilute samples, both P25 TiO2 and Cu2O/MTB suspensions show almost 100 % transmittances and appear similar to plain DI water. Interestingly, the two sets of transmittance spectra are quite different. For the P25 TiO2 suspensions, the transmittance spectra show a decreasing trend from 800 to 350 nm. However, the Cu2O/MTB suspensions exhibit almost flat spectra in transmittance, only slightly lower transmittances in the long-wavelength region, which may result from light scattering caused by the sub-micron size of Cu2O/MTB. The much lower transmittances of the P25 TiO2 suspensions in the low-wavelength region give rise to the light-blue colour of the suspensions as evident from the corresponding photos.

The results shown in Figure 6 suggest that the light shielding effect should be negligible when 0.0005 g of photocatalyst is suspended in 500 mL DI water. However, for suspensions that contain 0.01 and 0.05 g of photocatalyst, the light shielding effect should be strong enough to suppress the light utilisation. Here, we first investigate the H2 production efficiency of different photocatalysts in the absence of light shielding. the amount of H2 evolution produced by six different photocatalysts, which include P25 TiO2, commercial Cu2O powders, bare MTBs, and Cu2O/MTBs prepared at three different Cu2O precursor concentrations, are shown in Figure 7. All of the amounts of H2 evolution appear as linear functions of the reaction time, which indicates constant H2 evolution rates. If examined closely, however, the data points collected at the end of the first hour fall below where they should appear on the lines. This is caused by the necessary warm-up period of approximately 15 min for the light source to reach steady conditions. This warm-up period leads to slightly lower amounts of H2 production than expected.

thumbnail image

Figure 7. H2 evolution of commercial P25 TiO2 and Cu2O/MTB composites prepared at three different Cu2O precursor concentrations. 0.0005 g of photocatalyst was used for the measurements.

Download figure to PowerPoint

The results for the H2 evolution rates determined at the end of the first hour and at the end of the fifth hour are summarised in Table 3. Evidently, the H2 evolution rates determined based on the data for the first hour are significantly lower than those based on the data for the fifth hour. The results obtained from commercial P25 TiO2 are included for comparison. For P25 TiO2, the H2 evolution rate is estimated to be 44.9 μmol h−1 from the data for the fifth hour. The Cu2O/MTB composite, prepared with 0.1 M Cu2O precursor, gives a much higher H2 evolution rate of 197 μmol h−1. If we increase the Cu2O precursor concentration to 0.2 m, the H2 evolution rate is further increased to 223 μmol h−1, which is almost five times greater than that of P25 TiO2. However, if we further increase the precursor concentration to 0.3 m, the evolution rate decreases to 199 μmol h−1. For Cu2O/MTBs, the photoinduced electrons mainly come from the MTBs. Nevertheless, the Cu2O nanocrystals also contribute although to a minor extent. For TiO2 materials, charge transport plays a key role in photochemical applications, which include photocatalytic water splitting, photocatalytic degradation and photovoltaics. In photocatalytic water splitting, charge separation is a key factor to enhance the conversion efficiency. Consequently, sacrificial reagents are commonly used to scavenge electrons or holes to improve the charge separation. In this work, the Cu2O nanocrystals can also help improve charge separation through the matched band structure for electron–hole separation. It is evident from the data given in Table 3 that all three Cu2O/MTB samples show significantly higher (4.4–5-fold) H2 evolution rates than P25 TiO2. However, if the concentration of the Cu2O precursor is too high, too many Cu2O nanocrystals may be formed, which block the surface pores of the MTBs, reduce the surface area available for the photocatalytic water splitting events, prevent the sacrificial reagent from reaching the TiO2 surfaces for hole scavenging and lead to a lower improvement in H2 evolution rates. Consequently, there is an optimal Cu2O loading to give the highest H2 evolution rate.

Table 3. H2 evolution rates of P25 TiO2 and Cu2O/MTB composites prepared at three different Cu2O precursor concentrations.
0.0005 g sampleH2 evolution rate afterH2 evolution rate after
the first hour [μmol h−1]the fifth hour [μmol h−1]
P2523.744.9
Cu2O/MTB-0.1 M161197
Cu2O/MTB-0.2 M221223
Cu2O/MTB-0.3 M126199

We used photoluminescence (PL) spectroscopy to analyse the extent of the charge separation of the photocatalysts, and the results are shown in Figure 8 a. There are two main emission peaks observed for the TiO2-based samples, an excitonic emission at around 400–410 nm and a defect-induced one at around 465 nm.38 After the decoration of Cu2O nanocrystals, the intensities of both excitonic and defect-induced emissions are drastically reduced, and Cu2O/MTB-0.2 M shows the most quenched PL spectrum.39, 40 The quenching in PL is a good measure for the extent of electron–hole separation. Electron–hole separation is detrimental to PL emissions, which require effective charge re-combination, but beneficial for photocatalytic reactions. As expected, the extent of PL quenching correlates well with the H2 evolution rates summarised in Table 3.

thumbnail image

Figure 8. a) PL spectra of MTBs and Cu2O/MTB composites prepared at three different Cu2O precursor concentrations. b) Band structures of anataseTiO2 and Cu2O.

Download figure to PowerPoint

The band structures of Cu2O and anatase TiO2 are shown in Figure 8 b.41, 42 For Cu2O/MTBs, it is evident that the photoinduced electrons of Cu2O have a tendency to transfer to the conduction band of the MTBs, whereas the photoinduced holes of TiO2 have strong tendency to move to the valence band of Cu2O. The end result is well-separated electrons and holes. Notably, the holes at the valence band of Cu2O have a potential positive enough to oxidise the sacrificial reagent, methanol, the oxidation potential of which is smaller than 0.4 eV.43

As already demonstrated, light shielding will become pronounced at increased photocatalyst concentrations and leads to lower than expected H2 evolution rates. The H2 evolution curves for photocatalyst loadings of 0.0005, 0.01 and 0.05 g are shown in Figure 9. It is evident that, except for the case of 0.0005 g photocatalyst loading, the curves exhibit a saturating tendency, which is worse for the higher photocatalyst loading. First, let us examine H2 evolution rate over the first hour to study the effect of light shielding on H2 evolution. The H2 evolution rates for the first hour are 221, 2311 and 6644 μmol h−1 for photocatalyst loadings of 0.0005, 0.01 and 0.05 g, respectively. Ideally, if light shielding is absent from the system, the H2 evolution rate should correlate linearly with the amount of photocatalyst used in the system and should reach 4420 and 22 100 μmol h−1 for loadings of 0.01 and 0.05 g, respectively. In fact, the two catalysts only attain 52 (for 0.01 g) and 30 % (for 0.05 g) of the level of the 0.0005 g case. Evidently, the system with the highest photocatalyst concentration experiences more severe light shielding.

thumbnail image

Figure 9. H2 evolution for different catalyst loadings of Cu2O/MTBs prepared at a Cu2O precursor concentration of 0.2 M.

Download figure to PowerPoint

As for the saturating tendency, it is not caused by the deactivation of the photocatalysts but by the insufficient capacity of the reactor to accommodate the large amount of H2 generated from the water splitting. The partial pressure of H2 in the reactor increases, which leads to increased H2 solubility in the solution. The end result is the suppression of the H2 evolution reaction in the solution to give retarded H2 evolution.

To prove our point, a cycling test was conducted with a 1 h duration for the case of 0.05 g photocatalyst loading, and the results are shown in Figure 10. For the first cycle, the H2 production rate reaches an ultra-high value of 6644 μmol h−1 and maintains that high level for the second cycle. For the third and fourth cycles, the H2 production decreases slightly. This slight decrease in H2 evolution is caused by the trapping of a minor amount of photocatalyst in the reactor walls, which is not well dispersed in the solution for effective water splitting. In conclusion, the cycling test confirms that the photocatalyst can maintain its photocatalytic activity for a long period of time. The decreases in H2 evolution rates in long-term operation and cycling tests come from the limitation of the system hardware instead of the deactivation of the photocatalyst.

thumbnail image

Figure 10. Cycling H2 evolution for a loading of 0.05 g Cu2O/MTBs composite prepared at a Cu2O precursor concentration of 0.2 M.

Download figure to PowerPoint

A comparison of the performance of P25 TiO2 with the optimally decorated MTBs is shown in Figure 11 and the data are summarised in Table 4. Here, the H2 evolution data are taken at the end of the first hour to better avoid the hardware limitation issue. If the photocatalyst loading is 0.0005 g, both solutions appear limpid and the light shielding effect is negligible. The differences in H2 evolution come from the intrinsic properties of the photocatalysts. The H2 evolution of the Cu2O/MTBs is 9.3 times that of P25 TiO2. If the photocatalyst loading is 0.01 g, the transmittance of the Cu2O/MTB suspension at 350 nm is 50 % more than that of the P25 TiO2 suspension. The H2 evolution of the Cu2O/MTB suspension becomes 53 times that of the corresponding P25 TiO2 suspension. With a further increase in the amount of the photocatalyst to 0.05 g, the differences in transmittances between the two samples decrease and the performance ratio drops to 28-fold. From the above results, one may infer that the H2 evolution performance of the photocatalyst is affected significantly by the transmittances at low wavelengths, such as 350 nm, which are able to optically excite the TiO2-based materials for water splitting. If 0.0005 g of optimally decorated MTBs is used as the photocatalyst, an ultra-high specific H2 production rate of 223 mmol h−1 g−1 is achieved. However, if 0.05 g Cu2O/MTB is used as the photocatalyst, a high H2 production rate of 6.644 mmol h−1, which corresponds to a specific H2 evolution rate of 133 mmol h−1 g−1, can be obtained.

thumbnail image

Figure 11. Comparison of P25 TiO2 and Cu2O/MTBs at different loading amounts of the photocatalyst. The MTBs are decorated at a Cu2O precursor concentration of 0.2 M.

Download figure to PowerPoint

Table 4. H2 evolution rates and transmittances for suspensions that contain different amounts of P25 and Cu2O/MTB composites prepared at a Cu2O precursor concentration of 0.2 M.
Sample0.0005 g0.01 g0.05 g
H2T350 nmH2T350 nmH2T350 nm
[μmol h−1][%][μmol h−1][%][μmol h−1][%]
P2523.7≈10043.231.32400.6
Cu2O/MTB-0.2 M221≈100231181.6664422.1

Conclusions

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

Mesoporous TiO2 beads (MTBs), produced by a metal-salt-based hydro-thermal process, are successfully decorated with Cu2O nanocrystals by a simple chemical bath deposition process. The structural advantages of high specific surface areas, large pore volumes and suitable pore sizes of the MTBs in combination with the excellent charge separation realised through the decoration of band-structure-matched Cu2O nanocrystals, achieve an ultra-high specific H2 evolution rate of 223 mmol h−1 g−1 in the absence of light shielding and a large H2 evolution rate of 6.644 mmol h−1 at a photocatalyst loading of 0.05 g.

Experimental Section

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

Preparation of photocatalysts

The mesoporous TiO2 beads were produced by a metal-salt-based hydro-thermal process.34 In a typical run, ammonium chloride (0.267 g) and DI water (80 mL) were added into an autoclave with a capacity of 100 mL. The complete dissolution of the ammonium chloride was ensured before titanium sulfate solution (1 g; 24 %, Kanto Chemical) was introduced into the above solution. The autoclave was then sealed and heated to 200 °C for 12 h for the hydro-thermal treatment. After the hydro-thermal treatment, the precipitate was collected and rinsed with DI water twice and ethanol once by ultra-sonication and centrifugation before it was dried in an oven at 80 °C to afford the micron-sized MTBs.

A chemical bath deposition process was developed to decorate Cu2O nanocrystals onto the surfaces of the constituent nanocrystals of the MTBs. Typically, as-prepared MTBs (0.12 g) were added into an ethanolic solution of Cu(CH3COO)2 of a desired concentration. Ultra-sonication was then applied for 1 h to disperse the powders and allow the adsorption of Cu ions onto the surfaces of the constituent nanocrystals of the MTB. After the ion adsorption, the powder was collected by centrifugation and put into a NaOH solution (5 mL, 0.2 m). Ultra-sonication was again applied to disperse the powders. The well-dispersed suspension was then maintained at 60 °C in a water bath. Finally, an aqueous solution of glucose (5 mL, 0.2 m) was slowly added into the above suspension with constant stirring for a reaction time of 10 min. The precipitate was rinsed twice with DI water before it was dried overnight in an oven at 80 °C to afford the Cu2O/MTB composites.

H2 evolution measurements

For the photocatalytic reaction, a reactor equipped with an inner irradiation component was designed and used to efficiently use the light. The reaction temperature was maintained at 20 °C by using a circulatory water jacket. A 400 W high-pressure mercury lamp (HL400EH-5, SEN-LIGHTS) was used as the light source. An appropriate amount of photocatalyst was suspended in DI water (400 mL) as the sample for H2 evolution measurements. In addition to Cu2O/MTBs, commercial P25 TiO2 powders were also investigated for comparison. In addition, methanol (100 mL) was added as the sacrificial reagent to promote electron–hole separation and thus H2 evolution. The suspension was kept uniform with magnetic stirring. Prior to measurements, the reactor was purged with N2 for 15 min to remove the residual air and to ensure the air-tightness of the reactor. The H2 produced from the photocatalytic water splitting reaction was transported by a circulatory pump to a GC system (9800, China Chromatography) for detection, using Ar as the carrier gas.

Characterisation of photocatalysts

A field-emission scanning electron microscope (FESEM, S-4700, Hitachi) was used to observe the morphology and nanostructure of the photocatalysts. An X-ray powder diffractometer (XRD, MXP 18, MAC Science) was used to determine the crystalline phase and phase purity of the photocatalysts. The elemental composition and oxidation states of the photocatalyst surface were analysed by using a high-resolution X-ray photoelectron spectrometer (HRXPS, PHI Quantera, ULVAC-PHI). The micro-structural characteristics of the samples were determined by N2 adsorption–desorption analysis conducted at 77 K (NOVA e1000, Quantachrome). A UV/Vis spectrophotometer (U-3300, Hitachi) was used to measure the transmittance spectra of photocatalyst suspensions of different concentrations and a reflective UV/Vis spectrophotometer was used to measure the absorption spectra of the photocatalysts (UV-2450, Shimadzu). The optical emission characteristics of the photocatalysts were analysed by using a PL spectrophotometer (F-4500, Hitachi) with an excitation wavelength of 330 nm.

Acknowledgements

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

This work was financially supported by the National Science Council of the Republic of China (Taiwan) under grant NSC 101-2221-E-007-111-MY3 and by the Low Carbon Energy Research Center of the National Tsing-Hua University.