Preparation of layered titanate with interlayer cadmium sulfide particles for visible-light-assisted dye degradation

Abstract

Composite materials consisting of layered titanate and cadmium sulfide (CdS) particles were prepared via consecutive ion exchange and precipitation processes. Samples with different CdS loadings were characterized using X-ray diffraction, physical adsorption/desorption of nitrogen, scanning/transmission electron microscopy and ultraviolet-visible diffuse reflectance spectroscopy techniques. The photocatalytic properties of the samples in degrading methylene blue (MB) under visible light irradiation were evaluated. Results showed that CdS nanoparticles (NPs) were intercalated between titanate layers. The composite materials showed an improved photocatalytic activity as compared to both bulk CdS and a mixture of CdS and titanate. The intimate contact between titanate and CdS particles benefited interparticle electronic coupling. The observed improvement in photocatalytic activity is a combination of a number of contributions, including CdS photocatalysis, photosensitized degradation of MB over K₂Ti₄O₉, and the synergistic effect between CdS and K₂Ti₄O₉. The synergy indeed contributed to MB degradation and was found to depend on the content of CdS in the composite with an optimized content of 19 wt%, but it is not a dominant factor determining the overall photocatalytic activity of the composite.

---

Introduction

Heterogeneous photocatalysis has been explored for many applications, such as water splitting,^1–3 pollutant destruction,^4–7 and disinfection.^8–10 Many semiconducting photocatalytic materials have been developed so far.^11–20 Among them, cadmium sulfide (CdS) has a narrow band gap (∼2.52 eV), suitable for visible light photocatalysis.²¹ In addition, the edge of the conduction band (CB) of CdS is more negative than that of most semiconductors, favouring the production of reactive oxidative species (ROS) for advanced oxidation processes (AOP),¹² as well as for the reduction of the H₂O/H₂ redox potential for hydrogen evolution.^22,23 However, like other semiconductors, rapid electron–hole recombination occurs in CdS,^24–27 which hinders its application as a photocatalyst. Another problem of CdS is its poor stability against photocorrosion due to self-oxidation by photogenerated holes in the valence band (VB).¹

To solve the problems associated with CdS photocatalyst, embedding CdS in a solid matrix has been demonstrated to be a promising approach.^28–31 Chen et al.³⁰ showed that CdS-implanted layered double hydroxide (LDH) displayed an excellent performance in the degradation of rhodamine B under both UV and visible light irradiations. Shangguan and co-workers³¹ reported that CdS-intercalated metal oxides were superior to both CdS and a mixture of CdS and the metal oxide with respect to photocatalytic activity for hydrogen evolution. We also observed³² an enhanced photocatalytic performance of a CdS–graphene composite in the degradation of non-biodegradable dyes under visible light irradiation. The improvement is generally attributed to the fast transfer of photogenerated electrons from CdS to the second material, such as metal oxide or graphene.¹ Such a synergistic effect between CdS and the second material has been interpreted by comparing their band potentials.¹² However, due to the complicated nature of photocatalytic reactions,³³ they require an understanding of the role of each component in the composite including CdS, the second semiconducting material, the CdS/support interface and the interaction of these with chemical reactants during photocatalytic processes. To what extent does the synergistic effect between the two components determine the overall photocatalytic performance also needs further analysis? The study could have significant impact on understanding the underlying mechanism and designing advanced composite photocatalytic materials.

Here, we report the preparation of CdS–pillared titanate composite materials and their photocatalytic properties in degradation of methylene blue (MB) under visible light irradiation. Titanate was selected to couple with CdS because of the following reasons. First, the layered structure of titanate³⁴ could stabilize CdS particles. Second, titanate can be prepared as large particles,³⁵ making it easy to separate from the reaction mixture by for example sedimentation or filtration. Third, as has been demonstrated previously,¹¹ titanate possesses an excellent ability of accepting photogenerated electrons, thus having a potential to minimize the electron–hole recombination.³⁶ In the present work, titanate–CdS composites (designated as K₂Ti₄O₉–CdS) with different loadings of CdS were prepared and characterized. The role of each semiconductor component in the photocatalytic degradation of MB and the effect of CdS content on the synergistic contribution were investigated.

Experimental details

Synthesis of CdS–pillared titanate composites

Potassium tetratitanate (K₂Ti₄O₉) was synthesized according to the method described previously.¹¹ Potassium carbonate (Sigma-Aldrich) and TiO₂ (Sigma-Aldrich) with a molar ratio of 1 : 3 were ground and mixed. The mixture was heated at 150 °C for 2 h followed by annealing at 960 °C for 10 h with a heating rate of 15 °C min. The composition of K₂Ti₄O₉ was confirmed by X-ray diffraction (see Fig. 1).

Fig. 1 XRD patterns of different samples. The standard XRD pattern of CdS is included for reference. The data on the top left are the basal spacing in nanometres.

In the preparation of CdS–titanate composite materials, Cd²⁺ ions were firstly loaded onto K₂Ti₄O₉ via an ion exchange process, followed by precipitation in the presence of S²⁻ as illustrated in Scheme 1. Briefly, 50 mL of K₂Ti₄O₉ suspension (20 g L⁻¹) was mixed with 20 mL of Cd(NO₃)₂ (Sigma-Aldrich) solutions with different concentrations (0.24 M for K₂Ti₄O₉–CdS-1 and 0.81 M for K₂Ti₄O₉–CdS-2) under stirring. The mixture was stirred at 60 °C for 24 h. Then, a Na₂S solution (0.49 M, Sigma-Aldrich) was added dropwise. After continuous stirring at 60 °C for 3 h, yellow precipitates were collected by filtration, washed several times with deionized water, and dried in an oven at 70 °C. The samples are denoted as K₂Ti₄O₉–CdSx%, where x stands for the weight percentage of CdS in a composite. The content of CdS was determined by the energy dispersive X-ray (EDS) analysis.

Scheme 1 Schematic illustration of the preparation of CdS–pillared titanate.

For comparison purpose, a mixture containing K₂Ti₄O₉ and 19 wt% of CdS was prepared. This sample is denoted as (K₂Ti₄O₉–CdS)_mix.

Evaluation of photocatalytic activities and stability

The photocatalytic degradation of MB was carried out in an open thermostatic photoreactor (Fig. S1†). Before visible light irradiation, a suspension containing 100 mL of 6.1 × 10⁻⁵ M MB (Sigma-Aldrich) and 0.05 g of solid catalyst was sonicated for 5 min, followed by stirring for 30 min in the dark to allow sorption equilibrium. Then, the mixture was irradiated with a 350 W xenon lamp equipped with a 420 nm cut-off filter. At a given time interval, 5 mL aliquots were withdrawn using a syringe and filtered through a membrane (0.45 μm). The residual concentration of MB in the aliquots was analyzed using a Shimadzu 1601 PC UV-Vis spectrophotometer. It is noted that the reaction was conducted under weak acidic conditions (pH ∼ 6.3), which would enhance the protonation of MB and its adsorption on catalyst surface for photocatalytic reactions.³⁷ Each evaluation was carried out in triplicate and the data were expressed as means.

The stability of the photocatalyst was evaluated as follows. 0.05 g of K₂Ti₄O₉–CdS47% was suspended in 100 mL of 2.0 × 10⁻⁵ M MB solution. After a given time interval, 5 mL aliquots were withdrawn, filtered and analyzed on the UV-Vis spectrophotometer. After MB was completely degraded, the suspension was further irradiated for 20 min. Then, 20 mL stock solution (1 × 10⁻⁴ M) of MB was supplied to restore the initial concentration of MB to 2.0 × 10⁻⁵ M. The suspension was equilibrated again for 30 min in the dark before being irradiated by visible light.

Characterization

X-ray diffraction (XRD) measurements were performed on a Shimadzu diffractometer (XRD-6000, Tokyo, Japan) operated in the reflection mode with Cu Kα radiation at a step size of 0.06° per second. X-ray photoelectron spectroscopy (XPS) measurements were conducted on a PHI-5300 ESCA spectrometer (Perkin-Elmer) with an energy analyser working in the pass energy mode at 35.75 eV. An Al Kα line was used as the excitation source. Diffuse-reflectance spectra were recorded on a Shimadzu 3100 UV-Vis-NIR spectrometer equipped with an integrating sphere ISR-3100 using BaSO₄ as the reference. Photoluminescence spectra were collected at room temperature on a PTI luminescence spectrometer with a solid accessory. The specific surface area of samples was calculated using the Brunauer–Emmett–Teller (BET) method from the adsorption isotherms of N₂ measured at 77 K on a Quantachrome NOVA instrument. The morphologies of samples were characterized by a field-emission scanning electron microscope (FESEM, JEOL, JSM-6700F) and a high-resolution transmission electron microscopy (HRTEM, JEOL-2100, 200 kV).

Results and discussion

Fig. 1 shows the XRD patterns of K₂Ti₄O₉, CdS, and two K₂Ti₄O₉–CdS composite samples. The sharp and narrow diffraction peaks for the titanate sample are in good accordance with the literature data,¹¹ and can be indexed to (200), (201), (004), (512) and (205) reflections, indicating the formation of K₂Ti₄O₉ phase (JCPDS no. 32-0861). As a member of the M₂O·nTiO₂ family (M = Na, K, Rb, Cs) with n = 4,³⁵ K₂Ti₄O₉ has a lepidocrocite-type layered structure consisting of exchangeable interlayer K⁺ ions and the [Ti₄O₉]²⁻ layer built up from edge- and corner-shared TiO₆ octahedra,³⁸ which are also the basic unit of anatase titania.^35,38,39 When the K₂Ti₄O₉ solid was combined with Cd(NO₃)₂ solution, Cd²⁺ ions replaced K⁺ ions via an ion exchange process (Scheme 1) as confirmed by the XPS and EDX results (see Table 1). With the addition of S²⁻, CdS particles formed in situ and simultaneously deposited on the surface of K₂Ti₄O₉. The XRD pattern of sample K₂Ti₄O₉–CdS-1 showed some broad peaks around 26.5, 44.0 and 52.2 degrees two theta, corresponding to the standard reflection peaks of crystalline CdS (JCPDS Card no. 10-0454). The crystalline phase of K₂Ti₄O₉ remained unchanged except for a slight broadening of the (200) peak and the presence of two new peaks at around 7.5 and 9.6 degrees two theta, indicating the expansion of the basal spacing from 0.86 nm to 1.20 and 0.96 nm due to the intercalation of CdS into the lamellar structure.^11,35 As the initial [Cd²⁺] increased from 0.24 M to 0.81 M, the (200) peak of K₂Ti₄O₉ in composite K₂Ti₄O₉–CdS-2 completely shifted to lower angles with the intensity of the (111) diffraction peak for CdS also increased, suggesting that all interlayers of K₂Ti₄O₉ were completely intercalated with CdS particles. However, the XRD pattern for sample (K₂Ti₄O₉–CdS)_mix appeared to be a simple superimposition of each component. The position of each diffraction peak was almost unchanged, indicating that the mixing process did not lead to intercalation of CdS particles into the K₂Ti₄O₉ interlayers.

Table 1 XPS and EDX analysis results of samples

Sample	Element	XPS^a (%)	EDX^a^,^b (%)	EDX^b^,^c (wt%)
a Atomic concentration.b Average of four different selected areas.c Weight percentage.
K2Ti4O9–CdS-1	Cd	12.0	5.4	22.7
	S	4.5	3.6	4.3
	Ti	3.1	13.9	24.9
	K	∼0	2.1	3.1
K2Ti4O9–CdS-2	Cd	19.2	12.4	42.0
	S	13.9	10.9	10.5
	Ti	6.0	9.8	14.1
	K	∼0	0.7	0.8

---

Fig. 2 shows the FESEM and HRTEM images of sample K₂Ti₄O₉ and K₂Ti₄O₉–CdS-1. As can be seen, titanate K₂Ti₄O₉ has a rodlike morphology with a diameter of approximately 500 nm and lengths up to several micrometres (Fig. 2A). The surface of sample K₂Ti₄O–CdS-1 was covered by small aggregates with sizes ranging from 10 nm to 500 nm (Fig. 2B). The EDX data (Fig. 2C and D) confirmed the presence of CdS in the composite. The HRTEM images (Fig. 2E–G) further confirmed the rodlike morphology of titanate K₂Ti₄O₉, as well as the presence of CdS aggregates deposited on the surface of the rod. The dark regions on the K₂Ti₄O₉ background were probably due to the intercalated CdS particles. The lattice fringe indicates the highly crystalline nature of CdS. The interplanar spacing of the adjacent lattice fringes was estimated to be 0.34 nm, attributed to the (111) plane of CdS.

Fig. 2 FESEM images of K₂Ti₄O₉ (A) and K₂Ti₄O₉–CdS-1 (B). SEM image (C), EDX spectrum (D), and HRTEM images (E–G) of K₂Ti₄O₉–CdS-1.

Table 1 shows the elemental contents in two composite samples. It can be seen that when the initial [Cd²⁺] increased from 0.24 M to 0.81 M, the atomic ratio of K/Ti decreased from 0.15 to 0.07 with the ratio of Cd/Ti contradictorily increased from 0.39 to 1.27, indicating the displacement of K⁺ by Cd²⁺ during the ion exchange process. The result is slightly different from that determined by XPS analysis (Fig. S2†), in which element K was unable to be detected, and the atomic percentage of Cd was 12.0% in composite K₂Ti₄O₉–CdS-1, much larger than 5.4% determined by EDX. As XPS is an effective tool for surface analysis, K⁺ ions on the external surface of K₂Ti₄O₉ is thus believed to be fully displaced by Cd²⁺ ions, with some Cd ions intercalating into the interlayers as evidenced by the XRD data (Fig. 1). When the content of CdS increased from sample K₂Ti₄O–CdS-1 to sample K₂Ti₄O–CdS-2, the atomic percentage of Cd on the surface increased about 1.6 times, while the overall Cd content determined by EDX increased about 2.3 times, suggesting that most Cd ions were intercalated into the interlayers of K₂Ti₄O₉, resulting in an expansion of basal spacing as confirmed by the XRD data (Fig. 1). It is also noted that the atomic concentration of Cd determined by XPS is much larger than that of S, but they are comparable in the EDX analysis. The results suggest that there still have some Cd²⁺ ions adsorbed on the surface of K₂Ti₄O₉ that cannot be precipitated by S²⁻ ions. The resulting weight percentages of CdS in composites K₂Ti₄O₉–CdS are thus calculated by using that of the element S (Table 1), and are determined to be 19 wt% and 47 wt% for sample K₂Ti₄O₉–CdS-1 and K₂Ti₄O₉–CdS-2, respectively.

For (K₂Ti₄O₉–CdS)_mix, the FESEM image (see Fig. S3†) indicates that most CdS particles existed as aggregates in the mixture instead of depositing on the surface of K₂Ti₄O₉, which is in good accordance with the XRD analysis (Fig. 1).

The nitrogen adsorption–desorption isotherms and the pore size distribution curves of samples K₂Ti₄O₉, K₂Ti₄O₉–CdS19% and K₂Ti₄O₉–CdS47% are shown in Fig. S4.† It is seen that the K₂Ti₄O₉–CdS composites displayed a type IV isotherm, with a H3-type hysteresis loop, indicating the aggregation of plate-like particles giving rise to slit-shaped pores.⁴⁰ The Brunauer–Emmett–Teller (BET) specific surface area of samples K₂Ti₄O₉–CdS19% and K₂Ti₄O₉–CdS47% were calculated to be about 33.8 and 77.0 m² g⁻¹, respectively. The BET surface area of sample K₂Ti₄O₉ was only 2.2 m² g⁻¹. Similar findings have also been demonstrated in iron oxide⁴¹ and zinc oxide⁴² pillared titanate. The pore size distribution curves computed from the desorption branches using the Barrett–Joyner–Halenda (BJH) model clearly showed a narrow distribution with the main peak centred at about 3.5 nm and a weak shoulder peak at around 1.8 nm for both composite samples. On the basis of the data of both the pore size distribution curves and the basal spacing of the composite materials, the formation of large pores was due to the stacking of titanate platelets and the small pores to the basal spacing of layered K₂Ti₄O₉ upon intercalation of CdS particles. The porous structure of the composite materials is believed to favour the uptake of organic pollutants from solution to the catalyst surface for subsequent reactions.

Fig. 3 shows the UV-visible reflectance spectra of samples K₂Ti₄O₉, CdS, K₂Ti₄O₉–CdS, and (K₂Ti₄O₉–CdS)_mix. Layered titanate K₂Ti₄O₉ as a typical photocatalyst has a strong absorption in the UV region with the absorption edge at around 351 nm.^11,35,43 Composite K₂Ti₄O₉–CdS however showed a strong absorption in the visible light region with the absorption edge at around 539 nm for K₂Ti₄O₉–CdS19% and 579 nm for K₂Ti₄O₉–CdS47%. The bandgap thus can be estimated to be about 2.30 eV and 2.14 eV, respectively. Compared to that of bulk CdS (∼2.05 eV), the CdS NPs in K₂Ti₄O₉–CdS displayed a larger bandgap.³¹ As the bandgap strongly depends on the crystallite dimension of a semiconductor,⁴ the bandgap increases and the band edge shifts to yield larger redox potentials when the crystallite dimension decreases. The result suggests that the crystallite size of CdS particles in composite K₂Ti₄O₉–CdS are smaller than that of bulk CdS,^4,44 probably due to the confining effect of the K₂Ti₄O₉ interlayers. Moreover, the absorption intensity of composite K₂Ti₄O₉–CdS19% was much lower than that of (K₂Ti₄O₉–CdS)_mix, even though both samples had similar content of CdS, attributing to the intercalation of CdS into K₂Ti₄O₉ interlayers and the low extinction coefficient of K₂Ti₄O₉ in the visible region (Fig. 3). Nonetheless, the loading of CdS enabled the composite K₂Ti₄O₉–CdS to be active under visible light irradiation, whereas the intimate contact between CdS and K₂Ti₄O₉ and the increased redox potential of CdS will benefit the interparticle electronic coupling and charge transfer.

Fig. 3 UV-visible reflectance spectra of (a) K₂Ti₄O₉, (b) CdS, (c) (K₂Ti₄O₉–CdS)_mix, (d) K₂Ti₄O₉–CdS19%, and K₂Ti₄O₉–CdS47% (e).

The photocatalytic performance of the samples in the degradation of methylene blue (MB) under visible light irradiation is shown in Fig. 4. It can be seen that MB was very stable in the absence of a photocatalyst, or in the presence of photocatalysts without visible light irradiation. The concentration of MB was however slightly decreased in the presence of K₂Ti₄O₉ or (K₂Ti₄O₉–CdS)_mix in the dark. This concentration decrease is probably due to ion exchange between cationic MB dye and K⁺ in K₂Ti₄O₉.⁴⁵ As K₂Ti₄O₉ is only photocatalytically active under UV light irradiation, MB underwent moderate degradation in the K₂Ti₄O₉–MB system under visible light irradiation. With regard to photocatalyst CdS, it displayed a significantly higher degradation rate than that of K₂Ti₄O₉. Composite K₂Ti₄O₉–CdS47% (or K₂Ti₄O₉–CdS-2) gives the highest degree of MB degradation under the visible light irradiation. The performance of K₂Ti₄O₉–CdS47% is even higher than that of P25, which can degrade dye pollutants under visible light via a photosensitization mechanism.¹¹ The absorption of MB in the visible light region gradually decreased with irradiation time (Fig. S5†), indicating the complete destruction of aromatic structures. The degradation was well-fitted with a pseudo-first-order kinetics model ln(C_t/C₀) = −kt, where C_t and C₀ are concentrations of MB at time t and 0 min, respectively, and k is the rate constant, which were calculated to be 1.75 × 10⁻², 1.47 × 10⁻² and 3.48 × 10⁻³ min⁻¹ for samples K₂Ti₄O₉–CdS47%, CdS and K₂Ti₄O₉, respectively. The k gradually decreased when the initial concentration of MB increased (Fig. S6†). It is also seen that the degradation rate of MB in the presence of composite K₂Ti₄O₉–CdS19% was slightly slower than that of K₂Ti₄O₉–CdS47%, but faster than that of (K₂Ti₄O₉–CdS)_mix with the rate constants of about 1.29 × 10⁻² min⁻¹ for K₂Ti₄O₉–CdS19% sample and 8.24 × 10⁻³ min⁻¹ for (K₂Ti₄O₉–CdS)_mix sample. To further identify the visible light activity of K₂Ti₄O₉–CdS, benzoic acid that has no absorption in the visible light region was used as a probe molecule, which underwent gradual degradation under visible light irradiation (Fig. S7†), but it was hardly degraded by P25 or K₂Ti₄O₉ under the same experimental conditions.

Fig. 4 Concentration profiles of MB in the absence of a photocatalyst (a), in the presence of K₂Ti₄O₉ (b), CdS (c), (K₂Ti₄O₉–CdS)_mix (d), K₂Ti₄O₉–CdS19% (e), K₂Ti₄O₉–CdS47% (f) and P25 (g) under visible light irradiation. Concentration profiles of MB in the presence of K₂Ti₄O₉ (h), (K₂Ti₄O₉–CdS)_mix (i), and K₂Ti₄O₉–CdS19% (j) in the dark. The initial concentration of MB was 6.1 × 10⁻⁵ M.

Fig. S8† shows the stability of photocatalyst K₂Ti₄O₉–CdS47%. It can be seen that the material possessed an excellent stability after four times of reuse. The slightly declined photocatalytic activity is believed to be due to the loss of catalyst during sampling for the analysis of MB concentration.⁸

Considering that both CdS and titanate K₂Ti₄O₉ are semiconductor photocatalysts,^28,35 and only CdS can be excited by visible light irradiation. After bandgap excitation, the electron on the VB of CdS was injected into its CB (Fig. 5), followed by reacting with dissolved oxygen to form ROS such as HO˙, O₂˙⁻, HOO˙ and H₂O₂.^4,12 The valence hole together with the ROS attacked MB present in both solution and on the catalyst surface, producing various N-demethylated products, such as thionine and formaldehyde,⁴⁶ which were further mineralized to CO₂ and H₂O under prolonged light irradiation.⁴ For photocatalyst K₂Ti₄O₉, it cannot be excited by visible light due to its wide bandgap energy, which can only absorb UV light (Fig. 3). However, as titanate is semiconductive,^35,47 its CB can accept electrons generated from excited dye molecules as has been observed previously.⁴⁵ The electron in the CB reacts with surface-adsorbed O₂ to produce ROS for dye degradation. Such a photosensitization-assisted degradation over K₂Ti₄O₉ process is different from semiconductor photocatalysis,^8,11,45,48 and has proven to be an effective route for the removal of non-biodegradable dye pollutants.⁴⁸

Fig. 5 The potential diagram of CdS, MB, and K₂Ti₄O₉.

The possibility of electron transfer from excited MB to K₂Ti₄O₉ can be estimated using the standard redox potentials versus normal hydrogen electrode (NHE) (E_NHE), which can be estimated from the work functions (E) using E_NHE (V) = −4.5 − E (eV).¹³ Considering that the redox potentials of MB and excited MB (MB*) versus NHE are 1.17 V and −0.69 V,⁴⁹ and the potential energies of the CB and VB of K₂Ti₄O₉ are −4.02 and −7.56 eV,¹¹ respectively, the electron transfer from excited MB to K₂Ti₄O₉ is thus energetically feasible (see Fig. 5), accounting for the observed moderate degradation of MB over K₂Ti₄O₉ (Fig. 4). Such a degradation process has also been observed in other semiconducting systems including TiO₂,⁴⁸ graphene,^13,50 SnO₂,⁵¹ etc. However, considering the redox potentials of CB (−3.78 eV) and VB (−5.84 eV) of CdS,³² the degradation of MB over semiconductor CdS via the photosensitization process is unlikely due to the mismatched redox potentials (Fig. 5). Takizawa and co-workers⁴⁶ compared the degradation of MB and rhodamine B (RhB) in the presence of CdS photocatalyst, and also concluded that the degradation of MB over CdS was merely due to the photocatalysis of CdS rather than the photosensitized electron transfer from excited MB to the CB of CdS.

For composite K₂Ti₄O₉–CdS, the MB degradation is much more complicated. The observed overall performance consists of at least three processes: CdS photocatalysis, the photosensitized degradation of MB over K₂Ti₄O₉, and the synergistic contribution of the two semiconductors. Coupling K₂Ti₄O₉ with CdS could alleviate the charge carrier recombination in CdS and realize a vectorial transfer of photogenerated charge carriers from one component to the other due to their unique structural properties.¹² First, K₂Ti₄O₉ is a semiconductor with an anisotropic structure,^11,43 indicating a low surface and bulk irregularities and facilitating a high transport mobility of charge carriers in the bulk phase.⁵² Second, the redox potential of the CB of K₂Ti₄O₉ is much higher than that of CdS (Fig. 5), thus thermodynamically favouring direct electron transfer from CdS to K₂Ti₄O₉. We have previously observed that titanate can indeed accept electrons in a titanate–anatase core–shell composite through an interparticle charge carrier transition mechanism.^11,36 The higher potential gap between CdS and titanate K₂Ti₄O₉ (Fig. 5) than that between anatase and K₂Ti₄O₉ ¹¹ could impose a stronger driving force, enabling the charge transfer from the CB of CdS to that of K₂Ti₄O₉. The photoluminescence spectra of CdS, K₂Ti₄O₉–CdS19% and K₂Ti₄O₉–CdS47% shown in Fig. S9† confirmed the charge transfer process. It is seen that CdS has an intense fluorescence emission maximum at 516 nm due to the electron–hole recombination. After being combined with K₂Ti₄O₉, the matched band potentials (Fig. 5) drives the charge transfer from one particle to its neighbour to form a spatial separation between electrons and holes, leading to decreased luminescent intensity of CdS in composite K₂Ti₄O₉–CdS. This retardation improved the utilization of excitons and enhanced the production of ROS for MB degradation (Fig. 5).

It is noted that the observed concentration decrease of MB in the presence of composite K₂Ti₄O₉–CdS may also be partially due to ion exchange between cationic MB and K⁺ in K₂Ti₄O₉. This contribution was however insignificant (Fig. 4), because most of the ion exchange sites of titanate had already been occupied by Cd²⁺ during the preparation process. While the removal of MB by ion exchange between MB and Cd²⁺ can be neglected as is evidenced by the minor change of the MB concentration in the presence of K₂Ti₄O₉–CdS19% in the dark (Fig. 4).

Fig. 6 shows the rate constants for MB degradation over photocatalysts with different CdS loadings. It is seen that coupling CdS with semiconductor K₂Ti₄O₉ indeed enhanced the degradation efficiency with the optimum amount of CdS at about 47 wt%. From Fig. 6 and Table 2 it is then possible to normalize the rate constant with respect to CdS loaded onto K₂Ti₄O₉. We will focus on the interparticle interaction between CdS and K₂Ti₄O₉ support as the contribution of synergistic effect to the overall degradation requires further analysis. It is clear from the last column of Table 2 that the normalized rate constant of sample K₂Ti₄O₉–CdS19% is larger than that of sample K₂Ti₄O₉–CdS47%, suggesting the exertion of larger photocatalytic activity of CdS NPs in sample K₂Ti₄O₉–CdS19%. This may be due to the fact that a low loading of CdS NPs at K₂Ti₄O₉ surface benefits the dispersion of CdS (Fig. 2B) and the contact between the two semiconductors, thus enhancing the interparticle charge separation and depressing the charge recombination in excited CdS NPs. On contrast, a high loading of CdS may result in severe aggregation of CdS particles (Fig. S10†), weakening the interparticle interaction and the charge separation between excited CdS and K₂Ti₄O₉. In addition, the high photoactivity of sample K₂Ti₄O₉–CdS47% as compared to K₂Ti₄O₉–CdS19% suggests that the synergistic effect between coupled semiconductors with matched band potentials is not always a dominant factor determining the overall photocatalytic performance. Other factors, such as CdS photocatalysis and the photosensitization-assisted degradation over K₂Ti₄O₉ as observed in this work should also be considered when designing novel composite photocatalysts for different applications.

Fig. 6 Effect of CdS weight percentage on the rate constants of MB degradation.

Table 2 Photocatalytic degradation of MB over CdS, K₂Ti₄O₉ and K₂Ti₄O₉–CdS with different wt% of CdS

Samples	EDX CdS/Ti^a	k (min^−1)	Normalized k^b
a Atomic ratio.b Rate constant normalized with respect to CdS content is obtained by dividing the rate constant k (column 3) by the CdS/Ti ratio (column 2).
CdS	—	1.47 × 10^−2	—
K2Ti4O9	—	2.55 × 10^−3	—
K2Ti4O9–CdS
19 wt%	0.52	1.29 × 10^−2	2.48 × 10^−2
47 wt%	2.22	1.75 × 10^−2	1.12 × 10^−2

---

Conclusions

A composite photocatalyst of titanate with CdS particles between the titanate layers prepared through ion exchange and precipitation displayed an enhanced performance in degrading MB under visible light irradiation in comparison with bulk CdS and mixture of CdS and K₂Ti₄O₉. The composite was also stable under the reaction conditions after four cycles. The improved photocatalytic performance is attributed to the combination of CdS photocatalysis, photosensitization-assisted degradation over K₂Ti₄O₉, and the synergistic effect between the two components. Normalization of the rate constants with CdS deposited at K₂Ti₄O₉ surface showed that the synergistic effect depends on the content of CdS, and is not a decisive factor determining the overall performance of the composite photocatalyst.

Acknowledgements

Australian Research Council (ARC) is acknowledged for supporting the work under the ARC Future Fellow Program (FT100100879). ZGX is grateful of the financial support from UQ new staff research start-up fund (2012000640).

Notes and references

1. X. Chen, S. Shen, L. Guo and S. S. Mao, Chem. Rev., 2010, 110, 6503–6570.
2. A. Kudo and Y. Miseki, Chem. Soc. Rev., 2009, 38, 253–278.
3. X. B. Chen, L. Liu, P. Y. Yu and S. S. Mao, Science, 2011, 331, 746–750.
4. M. R. Hoffmann, S. T. Martin, W. Y. Choi and D. W. Bahnemann, Chem. Rev., 1995, 95, 69–96.
5. H. G. Yang, C. H. Sun, S. Z. Qiao, J. Zou, G. Liu, S. C. Smith, H. M. Cheng and G. Q. Lu, Nature, 2008, 453, 638–641.
6. R. Asahi, T. Morikawa, T. Ohwaki, K. Aoki and Y. Taga, Science, 2001, 293, 269–271.
7. M. D'Arienzo, J. Carbajo, A. Bahamonde, M. Crippa, S. Polizzi, R. Scotti, L. Wahba and F. Morazzoni, J. Am. Chem. Soc., 2011, 133, 17652–17661.
8. Z. G. Xiong, J. Z. Ma, W. J. Ng, T. D. Waite and X. S. Zhao, Water Res., 2011, 45, 2095–2103.
9. J. H. Pan, H. Q. Dou, Z. G. Xiong, C. Xu, J. Z. Ma and X. S. Zhao, J. Mater. Chem., 2010, 20, 4512–4528.
10. A. Fujishima, X. T. Zhang and D. A. Tryk, Surf. Sci. Rep., 2008, 63, 515–582.
11. Z. G. Xiong and X. S. Zhao, J. Am. Chem. Soc., 2012, 134, 5754–5757.
12. H. J. Zhang, G. H. Chen and D. W. Bahnemann, J. Mater. Chem., 2009, 19, 5089–5121.
13. Z. G. Xiong, L. L. Zhang, J. Z. Ma and X. S. Zhao, Chem. Commun., 2010, 46, 6099–6101.
14. Z. G. Xiong, Y. M. Xu, L. Z. Zhu and J. C. Zhao, Environ. Sci. Technol., 2005, 39, 651–657.
15. A. H. Sun, Z. G. Xiong and Y. M. Xu, J. Mol. Catal. A: Chem., 2006, 259, 1–6.
16. C. Han, M. Pelaez, V. Likodimos, A. G. Kontos, P. Falaras, K. O'Shea and D. D. Dionysiou, Appl. Catal., B, 2011, 107, 77–87.
17. T. Tachikawa, M. Fujitsuka and T. Majima, J. Phys. Chem. C, 2007, 111, 5259–5275.
18. R. Nakamura, A. Okamoto, H. Osawa, H. Irie and K. Hashimoto, J. Am. Chem. Soc., 2007, 129, 9596–9597.
19. J. Zhang, Y. Wu, M. Xing, S. A. K. Leghari and S. Sajjad, Energy Environ. Sci., 2010, 3, 715–726.
20. S. Sajjad, S. A. K. Leghari and J. Zhang, RSC Adv., 2013, 3, 1363–1367.
21. Y. Liu, Y.-X. Yu and W.-D. Zhang, J. Alloys Compd., 2013, 569, 102–110.
22. W. Q. Cui, S. S. Ma, L. Liu, J. S. Hu and Y. H. Liang, J. Mol. Catal. A: Chem., 2012, 359, 35–41.
23. D. Jing and L. Guo, J. Phys. Chem. B, 2006, 110, 11139–11145.
24. Z.-D. Meng, M.-M. Peng, L. Zhu, W.-C. Oh and F.-J. Zhang, Appl. Catal., B, 2012, 113–114, 141–149.
25. L. J. Zhang, R. Zheng, S. Li, B. K. Liu, D. J. Wang, L. L. Wang and T. F. Xie, ACS Appl. Mater. Interfaces, 2014, 6, 13406–13412.
26. J. Zhang, Y. Wang, J. Jin, J. Zhang, Z. Lin, F. Huang and J. Yu, ACS Appl. Mater. Interfaces, 2013, 5, 10317–10324.
27. J. Zhang, S. Z. Qiao, L. Qi and J. Yu, Phys. Chem. Chem. Phys., 2013, 15, 12088–12094.
28. Q. Y. Li, T. Kako and J. H. Ye, J. Mater. Chem., 2010, 20, 10187–10192.
29. T. Sato, K. Masaki, K. I. Sato, Y. Fujishiro and A. Okuwaki, J. Chem. Technol. Biotechnol., 1996, 67, 339–344.
30. Y. Guo, H. Zhang, Y. Wang, Z.-L. Liao, G.-D. Li and J.-S. Chen, J. Phys. Chem. B, 2005, 109, 21602–21607.
31. W. Shangguan and A. Yoshida, J. Phys. Chem. B, 2002, 106, 12227–12230.
32. Z. Xiong, L. L. Zhang, J. Zhang and X. S. Zhao, Nanosci. Nanotechnol.--Asia, 2012, 2, 79–89.
33. Y. Liu, Y.-X. Yu and W.-D. Zhang, J. Phys. Chem. C, 2013, 117, 12949–12957.
34. T. W. Kim, S. G. Hur, S. J. Hwang, H. Park, W. Choi and J. H. Choy, Adv. Funct. Mater., 2007, 17, 307–314.
35. M. R. Allen, A. Thibert, E. M. Sabio, N. D. Browning, D. S. Larsen and F. E. Osterloh, Chem. Mater., 2010, 22, 1220–1228.
36. Z. G. Xiong and X. S. Zhao, J. Mater. Chem. A, 2013, 1, 7738–7744.
37. W. Cui, S. Ma, L. Liu, J. Hu, Y. Liang and J. G. McEvoy, Appl. Surf. Sci., 2013, 271, 171–181.
38. G. Liu, L. Z. Wang, C. H. Sun, X. X. Yan, X. W. Wang, Z. G. Chen, S. C. Smith, H. M. Cheng and G. Q. Lu, Chem. Mater., 2009, 21, 1266–1274.
39. H. Y. Zhu, X. P. Gao, Y. Lan, D. Y. Song, Y. X. Xi and J. C. Zhao, J. Am. Chem. Soc., 2004, 126, 8380–8381.
40. K. S. W. Sing, D. H. Everett, R. A. W. Haul, L. Moscou, R. A. Pierotti, J. Rouquerol and T. Siemieniewska, Pure Appl. Chem., 1985, 57, 603–619.
41. T. W. Kim, H. W. Ha, M. J. Paek, S. H. Hyun, I. H. Baek, J. H. Choy and S. J. Hwang, J. Phys. Chem. C, 2008, 112, 14853–14862.
42. K. Z. Zhang, B. Z. Lin, Y. L. Chen, B. H. Xu, X. T. Pian, J. D. Kuang and B. Li, J. Colloid Interface Sci., 2011, 358, 360–368.
43. N. Z. Bao, X. Feng, Z. H. Yang, L. M. Shen and X. H. Lu, Environ. Sci. Technol., 2004, 38, 2729–2736.
44. J. H. He, I. Ichinose, T. Kunitake and A. Nakao, Langmuir, 2002, 18, 10005–10010.
45. N. Miyamoto, K. Kuroda and M. Ogawa, J. Phys. Chem. B, 2004, 108, 4268–4274.
46. T. Takizawa, T. Watanabe and K. Honda, J. Phys. Chem., 1978, 82, 1391–1396.
47. S. F. Cheng, S. J. Tsai and Y. F. Lee, Catal. Today, 1995, 26, 87–96.
48. C. C. Chen, W. H. Ma and J. C. Zhao, Chem. Soc. Rev., 2010, 39, 4206–4219.
49. T. Shen, Z.-G. Zhao, Q. Yu and H.-J. Xu, J. Photochem. Photobiol., A, 1989, 47, 203–212.
50. Z. G. Xiong, L. L. Zhang and X. S. Zhao, Chem.–Eur. J., 2011, 17, 2428–2434.
51. J. T. Zhang, Z. G. Xiong and X. S. Zhao, J. Mater. Chem., 2011, 21, 3634–3640.
52. Z. G. Xiong, H. Wu, L. H. Zhang, Y. Gu and X. S. Zhao, J. Mater. Chem. A, 2014, 2, 9291–9297.

---

Footnote

† Electronic supplementary information (ESI) available: FESEM images of sample K₂Ti₄O₉–CdS47% and (K₂Ti₄O₉–CdS)_mix, N₂ sorption–desorption isotherms and pore size distribution curves, photoluminescence spectra of CdS and composite K₂Ti₄O₉–CdS, and the recycled degradation of MB. See DOI: 10.1039/c4ra09692d

---