Microwave-assisted synthesis of CdS intercalated K4Nb6O17 and its photocatalytic activity for hydrogen production

Applied Catalysis A: General 417–418 (2012) 111–118

Contents lists available at SciVerse ScienceDirect

Applied Catalysis A: General journal homepage: www.elsevier.com/locate/apcata

Microwave-assisted synthesis of CdS intercalated K4 Nb6 O17 and its photocatalytic activity for hydrogen production Wenquan Cui ∗ , Yanfei Liu, Li Liu, Jinshan Hu, Yinghua Liang College of Chemical Engineering, Hebei United University, Tangshan, 063009, PR China

a r t i c l e

i n f o

Article history: Received 19 October 2011 Received in revised form 15 December 2011 Accepted 18 December 2011 Available online 27 December 2011 Keywords: Microwave Intercalation CdS K4 Nb6 O17 Photocatalytisis, Hydrogen production

a b s t r a c t CdS intercalated K4 Nb6 O17 (designated as K4 Nb6 O17 /CdS) composite photocatalysts were synthesized using a microwave-assisted synthesis. The composite particles were characterized by X-ray diffraction (XRD), scanning electron microscopy (SEM), X-ray fluorescence spectrometry (XRF), energy dispersive Xray (EDS), ultraviolet–visible diffuse reflection spectra (UV–vis) and photoluminescence measurements (PL). The photocatalytic properties of these catalysts for hydrogen production were also investigated in the presence of Na2 S and Na2 SO3 sacrificial reagents. The K4 Nb6 O17 /CdS catalysts synthesized using microwave irradiation were found to possess higher crystallinity than their counterparts, synthesized using conventional methods. The absorption edge of K4 Nb6 O17 was shifted to the visible light region after the intercalation of CdS. Compared to the conventional synthesis method, the use of microwave irradiation sharply shortened the intercalation time. Furthermore, the K4 Nb6 O17 /CdS sample prepared via microwave irradiation exhibited higher activities for photocatalytic hydrogen production under both UV light and visible light irradiation, and the amounts of hydrogen produced were 265.95 mmol/(g cat) and 5.68 mmol/(g cat) after 3 h of irradiation, respectively. The mechanism of separation of the photogenerated electrons and holes at the K4 Nb6 O17 /CdS composite was discussed. © 2011 Elsevier B.V. All rights reserved.

1. Introduction The decomposition of water into hydrogen and oxygen by photoinduced processes employing a semiconductor irradiated with light can be used for sustainable solar energy conversion, and has been investigated in recent years [1–4]. Layered semiconductors are a group of compounds which demonstrate high activities for photocatalytic water splitting, and some of these semiconductors such as K4 Nb6 O17 [5], K2 Ti4 O9 [6] and K2 La2 Ti3 O10 [7] have been reported to facilitate hydrogen production using photo-processes. K4 Nb6 O17 is a two-dimensional layered compound, which is formed by the connection of octahedral units of NbO6 and oxygen bridging atoms. K4 Nb6 O17 , like other layered compounds, cannot exhibit any photocatalytic activity under visible light irradiation, due to its relatively large band gap (about 3.2 eV [8]). The intercalation of nano-semiconductor particles with narrow band gap in the layered niobates of K4 Nb6 O17 can extend the absorption edge into the visible light region, and hence improve the photocatalytic activity by using these narrow band gap semiconductors as sensitizers. Some researchers reported the successful improvement of photocatalytic activity of K4 Nb6 O17 under visible light through the intercalation of nano-sized CdS [9,10]. The existence of nano-sized

∗ Corresponding author. E-mail address: [email protected] (W. Cui). 0926-860X/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.apcata.2011.12.027

CdS particles in the layered space of K4 Nb6 O17 compounds cannot only prevent the photo-corrosion of these nano-particles, but can also promote the efficiency of separation of photogenerated elections and holes, facilitating the photocatalytic activity of hydrogen evolution under visible light irradiation. The preparation of CdS intercalated K4 Nb6 O17 via conventional methods can take several days to several weeks [9,11] due to mass transfer limitations. Additionally, the preparation process is difficult to control, and the crystal structure of K4 Nb6 O17 could be consequently destroyed [11,12]. It has been reported that increasing the crystallinity of nano-TiO2 could enhance photocatalytic activity [13], because of higher migration rates of photo generated carriers in the crystals. Additionally, methylthionine chloride intercalated K4 Nb6 O17 catalysts were prepared in the literature, and the results obtained indicated that the photocatalytic activity of the catalysts could be effectively improved by selecting K4 Nb6 O17 with higher crystallinity [14]. Microwaves consist of electromagnetic waves whose wavelength varies from 1 mm to 1 m, and application of this electric field causes the vibration of dipolar molecules. An electromagnetic field changes its direction at a frequency of hundreds of millions or even billions times per second. Since the dipolar molecules cannot keep up with the rate of electromagnetic field conversion, friction and the consequent production of heat ensues. In this sense, microwave energy is extremely efficient in the selective heating of materials. Compared to the conventional treatment, this has the effect of

112

W. Cui et al. / Applied Catalysis A: General 417–418 (2012) 111–118

sharply shortening the reaction time and increasing reaction efficiency [15]. Some intercalated materials, such as kaolinite/dimethyl sulfoxide [15], clays/Al [16], and C17 H35 COO− /Ni-Al-HTLc [17], etc., have been synthesized using the microwave irradiation. The microwave-assisted intercalation was found to not only decrease the reaction time, but also maintain a high crystallinity of the materials, suggesting that this method is advantageous over conventional means for the synthesis of intercalated compounds. At present, to our knowledge, there are no studies on the preparation of microwave-assisted intercalation of K4 Nb6 O17 and its photocatalytic activity for hydrogen production. In this paper, we prepared CdS intercalated K4 Nb6 O17 composite materials under microwave irradiation for the first time. The structure of the samples was characterized by XRD, SEM, X-ray fluorescence spectrometry, UV–vis and photoluminescence measurements. The photocatalytic activities of the samples synthesized via different methods were tested on hydrogen evolution in the presence of Na2 S and Na2 SO3 sacrificial reagent under UV light and visible light irradiation, respectively.

2. Experimental 2.1. Synthesis of photocatalysts All of the reagents were analytical grade and used without further purification. The K4 Nb6 O17 powder was obtained by high temperature solid phase reaction [18]. The detailed procedure is as follows: mixtures of K2 CO3 and Nb2 O5 (mol ratio 2.1:3), were ground in an agate mortar. The mixture was calcined at 1000 ◦ C for 2 h, and the product was further ground to a fine powder for subsequent use. The CdS intercalated K4 Nb6 O17 photocatalyst was synthesized under the assistance of microwave irradiation. The detailed procedure is as follows: first, the as-prepared K4 Nb6 O17 powder was added into 1 mol/L of HCl solution, and then stirred under microwave irradiation (400 W, 2.45 GHz) for 1 h in order to replace K+ with H+ in the layered space of K4 Nb6 O17 . The microwave irradiation was used as an intermittent process, in which the microwave was suspended for 2 min after 2 min of irradiation in order to avoid drastically boiling the solution. The temperature of the solution was maintained at about 85 ◦ C during the course of the reaction. After the microwave treatment, the suspension was centrifuged and washed using distilled water and then dried at 60 ◦ C in a far infrared drying oven for 10 h. The sample prepared in this manner was designated as m-H4 Nb6 O17 . Butylamine intercalated K4 Nb6 O17 was synthesized via the reaction of m-H4 Nb6 O17 with 50% volume ratio of butylamine in an aqueous solution under microwave irradiation at 55 ◦ C for 3 h. After that irradiation period, the sample was centrifuged, washed and dried. This material was designated as m-(C4 H9 NH3 )4 Nb6 O17 . To facilitate the ion-exchange, the m-(C4 H9 NH3 )4 Nb6 O17 sample was reacted with Cd(CH3COO)2 in a 0.4 mol/L aqueous solution under microwave heating at 70 ◦ C for 3 h. This suspension was centrifuged and washed carefully using distilled water until the pH value of the solution reached 7, indicating the removal of Cd2+ on the surface of K4 Nb6 O17 . The dried Cd intercalated K4 Nb6 O17 product was then reacted with H2 S in a U-tube, thereby intercalating CdS into the K4 Nb6 O17 . H2 S was obtained by the reaction of sulfuric acid and Na2 S. The catalyst was designated as mm-K4 Nb6 O17 /CdS. In this paper, we also prepared the following samples using conventional synthesis methods for comparison. n-H4 Nb6 O17 was prepared via stirring the mixture of K4 Nb6 O17 and 1 mol/L HCl solution using heating via a water bath at 25 ◦ C for 72 h; n-(C4 H9 NH3 )4 Nb6 O17 was prepared by butylamine pillared n-H4 Nb6 O17 and water bath heating at 60 ◦ C for 72 h.

mn-K4 Nb6 O17 /CdS was obtained via stirring the mixture of Cd(CH3COO)2 and m-(C4 H9 NH3 )4 Nb6 O17 in an aqueous solution using conventional water bath heating at 70 ◦ C for 5 h, followed by the sulfurization in a H2 S gas flow. The sample labeled nm-K4 Nb6 O17 /CdS was obtained via stirring the mixture of Cd(CH3COO)2 and n-(C4 H9 NH3 )4 Nb6 O17 under microwave irradiation. nn-K4 Nb6 O17 /CdS was synthesized conventionally by using water bath heating throughout the whole process. Pure CdS was synthesized by the reaction of the Cd(COOH)2 (1 mol/L) and Na2 S (1 mol/L) in an aqueous solution. M-CdS/K4 Nb6 O17 was prepared by the direct mechanical mixture of CdS and K4 Nb6 O17 powder in a quality ratio of 1:20. 2.2. Characterization of the photocatalysts The crystal structure and the phase of the samples were determined by X-ray diffractometer (XRD) using a Rigaku D/MAX2500 PC diffractometer with Cu K␣ radiation, with an operating voltage of 40 kV and an operating current of 100 mA. The average size in different crystal planes was estimated using the Scherrer’s equation. The morphology of the samples was detected using a scanning electron microscope (SEM) (Hitachi, s-4800). UV–visible light (UV–vis) diffuse reflectance spectra were recorded on a UV–vis spectrometer (Puxi, UV1901). The chemical compositions of the sample were tested by an energy dispersive X-ray detector (EDS, Thermo Noran 7) and an X-ray fluorescence spectrometer (Rigaku, ZSX PromusII). The luminescence of the powdered samples was measured on a spectrofluorometer (Hitachi, f7000). 2.3. Photocatalytic activity The photocatalytic activities of hydrogen production for the photocatalyst samples were examined in an inner irradiation system. Typically, 0.5 g powder of photocatalyst was dispersed in a Pyrex reaction cell containing 250 ml of aqueous solution of 0.1 mol/L Na2 S, 0.5 mol/L Na2 SO3 and 1 mol/L KOH. Thermostatic water flowed through a jacket between the light source and the reaction cell to remove extra heat produced by the lamp. The light sources used were a 300 W mercury lamp (UV) and a 500 W xenon lamp (the light with  < 400 nm was filtered out by flowing 1 mol/L NaNO2 solution in the jacket between the xenon lamp and the reaction cell). The suspensions were deairated with Ar gas for 30 min to prevent uptake of photo-generated electrons by dissolved oxygen before irradiation. The produced hydrogen gas was detected using an online gas chromatography system (FULI-GC9790, molecular sieve 5 A column, TCD detector, Ar carrier). The Ar gas was used as a carrier for the products. 3. Results and discussion 3.1. Characterization of catalysts Fig. 1 shows XRD patterns of the samples obtained from various stages of the preparation process of CdS intercalated K4 Nb6 O17 . As shown in Fig. 1(A), the XRD peaks of potassium niobium (curve a) can be readily indexed to a pure phase of orthorhombic K4 Nb6 O17 , according to JCPDS (31-1064). The diffraction peaks are intense and sharp, indicating that the obtained K4 Nb6 O17 is well crystallized. The main peak appears at 10.68◦ , corresponding to the 0 4 0 crystal plane, indicating the d value is 0.83 nm. The width of the layered space is 0.27 nm and was calculated by subtracting the Nb6 O17 4+ layer thicknesses of 0.56 nm [19] from the d0 4 0 value. Curve b in Fig. 1(A) shows the XRD pattern of m-H4 Nb6 O17 prepared via ion exchange under microwave irradiation. The main peak of the 0 4 0 crystal face is at 11.02◦ , indicating the width of the layered

W. Cui et al. / Applied Catalysis A: General 417–418 (2012) 111–118

113

A d

040

c

Intensity /a.u

040

b

040 040

0

a

10

20

30

40

50

60

o 2θ/( )

B 140

040

0160

140

200 032 132 200 0100

040 140

10

0100 132

c 0160

040

0

d

0100 140

Intensity /a.u

C

200 0100 032 132

b

a 30

2θ/(

040

140 200

040

140

200

040

040

132

200 0100

20

Intensity /a.u

040

40

50

60

o

)

200

d

0100 132

0100

c

132

140

140

10

200

20

0100 132

30 2θ/(

40

b

a

50

60

o)

Fig. 1. XRD spectra of samples. (A) ((a) K4 Nb6 O17 (b) m-H4 Nb6 O17 (c) m-(C4 H9 NH3 )4 Nb6 O17 (d) mm-K4 Nb6 O17 /CdS); (B) ((a) n- H4 Nb6 O17 (b) m-H4 Nb6 O17 (c) n(C4 H9 NH3 )4 Nb6 O17 (d) m-(C4 H9 NH3 )4 Nb6 O17 ); (C) ((a) nn-K4 Nb6 O17 /CdS (b) nm-K4 Nb6 O17 /CdS (c) mn-K4 Nb6 O17 /CdS (d) mm-K4 Nb6 O17 /CdS).

space of m-H4 Nb6 O17 is 0.24 nm, which is narrower than that of K4 Nb6 O17 . This change can be explained by the replacement of K+ with H+ , because the radius of H+ is 0.032 nm, smaller than that of K+ (0.133 nm). Therefore, the width of the space decreased with the substitution of K+ by H+ . Additionally, the results show evidence of successful entrance of H+ into the interlayer of K4 Nb6 O17 using microwave irradiation. Curve c is the XRD pattern of m(C4 H9 NH3 )4 Nb6 O17 . As shown, the main peak (0 4 0) shifted to 5◦ , which corresponds with the increase of the width of the layered space of m-(C4 H9 NH3 )4 Nb6 O17 . Curve d is the XRD pattern of the mm-K4 Nb6 O17 /CdS. Compared to m-(C4 H9 NH3 )4 Nb6 O17 , the peak at about 8.140◦ is assigned to the 040 crystal plane, indicating a decrease of the layered space. These results confirmed that CdS was successfully intercalated into the interlayer space of K4 Nb6 O17 under the aid of microwave irradiation. The XRD patterns of n-H4 Nb6 O17 , m-H4 Nb6 O17 , n(C4 H9 NH3 )4 Nb6 O17 and m-(C4 H9 NH3 )4 Nb6 O17 are shown in Fig. 1(B). The diffraction peaks of n-H4 Nb6 O17 corresponding to the (1 4 0), (2 0 0), (0 10 0) and (1 3 2) crystal planes were obviously weaker than that of m-H4 Nb6 O17 . The diffraction peaks of n-H4 Nb6 O17 corresponding to (0 3 2) and (1 16 0) crystal planes disappeared. A reasonable suggestion from this observation is that m-H4 Nb6 O17 had a higher crystallinity than n-H4 Nb6 O17 . The peaks corresponding to 0 4 0 crystal plane for m-(C4 H9 NH3 )4 Nb6 O17 and m-H4 Nb6 O17 were stronger than those of n-(C4 H9 NH3 )4 Nb6 O17 and n-H4 Nb6 O17 , respectively. This suggested that microwave irradiation could improve the efficiency of the H+ exchange and butylamine intercalation procedures. The strong (0 4 0) diffraction

peak indicated well ordered stacking of niobate sheets [20], so the higher efficiency of H+ exchange and butylamine intercalation could result in the larger interlayer space and weaker connection between niobate sheets, leading to the weak diffraction peak observed for the 0 4 0 crystal plane. As such, the high efficiency of H+ exchange and butylamine intercalation can improve intercalation of CdS, which is consistent with the XRF data. The appearance of a new diffraction peak at 9.34◦ in n-H4 Nb6 O17 in Fig. 1(B) could be assigned to the intercalation of H2 O. The average sizes of different crystal planes were calculated using the Scherrer equation. The particle size of K4 Nb6 O17 , mH4 Nb6 O17 , and n-H4 Nb6 O17 were found to be 59.8 nm, 31.8 nm, and 45.4 nm, respectively. The size of m-H4 Nb6 O17 particle was smaller than that of K4 Nb6 O17 and n-H4 Nb6 O17 , indicating that the use of microwave irradiation could reduce the particle size obtained. Fig. 1(B) also shows the XRD patterns of n-(C4 H9 NH3 )4 Nb6 O17 prepared using conventional synthesis and m-(C4 H9 NH3 )4 Nb6 O17 prepared under microwave irradiation, respectively. The m-(C4 H9 NH3 )4 Nb6 O17 sample has a similar structure as n(C4 H9 NH3 )4 Nb6 O17 , however the m-(C4 H9 NH3 )4 Nb6 O17 showed obvious peaks assignable to the (2 0 0), (0 3 2) and (1 3 2) crystal planes. The XRD patterns of nn-K4 Nb6 O17 /CdS, nm-K4 Nb6 O17 /CdS, mnK4 Nb6 O17 /CdS and mm-K4 Nb6 O17 /CdS are given in Fig. 1(C). As shown, the samples exhibited similar diffraction patterns. This indicated that there are negligible differences during the Cd2+ intercalation process when using the traditional synthesis method and microwave irradiation, respectively. The peaks of crystal plane 0 4 0

114

W. Cui et al. / Applied Catalysis A: General 417–418 (2012) 111–118

Fig. 2. SEM images of photocatalysts ((a) K4 Nb6 O17 ; (b) mm-K4 Nb6 O17 /CdS, (c) CdS; (d) M-K4 Nb6 O17 /CdS); (e) corresponding EDS pattern for (b).

of nn-K4 Nb6 O17 /CdS and nm-K4 Nb6 O17 /CdS are both at 9.5◦ , but the peaks of mm-K4 Nb6 O17 /CdS and mn-K4 Nb6 O17 /CdS shift to 8.14◦ , which could suggest that the interlayer spacing increased and there was more CdS entering into the K4 Nb6 O17 interlayer. The morphologies of the samples were observed using SEM, shown in Fig. 2. Fig. 2(a) presents a typical SEM image of the pure K4 Nb6 O17 , which showed aggregated sheet-like morphology with the stacks of layered structures. From Fig. 1(a), the K4 Nb6 O17 particles exhibited relatively uniform size distribution and smooth surface morphology, with a size of approximately 2–4 ␮m. The morphology of mm-K4 Nb6 O17 /CdS, shown in Fig. 2(b), was found to be similar to that of pure K4 Nb6 O17 , suggesting that the shape of K2 Ti4 O9 did not change. This indicated that the microstructure of the sample did not change drastically after CdS intercalation under microwave irradiation. Fig. 2(c) is the image of pure CdS, showing agglomerated particles, with size varying from 0.5 ␮m to 4 ␮m or even more, indicating that the precipitation resulted in large and nonuniform of pure CdS. Fig. 2(d) is the image of M-CdS/K4 Nb6 O17 , which shows that CdS and K4 Nb6 O17 are directly blended together. Fig. 2(e) shows the typical EDS spectrum obtained from mmK4 Nb6 O17 /CdS. In the spectrum, peaks associated with O, S, Cd, K, and Nb were observed. K, Nb, and O peaks result from K4 Nb6 O17 , and Cd and S result from CdS, respectively. The EDS results

confirmed that the obtained product is composed of the K4 Nb6 O17 intercalated with CdS. To further explore the chemical composition XRF analysis was also employed to determine the elemental content of the asprepared CdS samples, as shown in Table 1. The existence of K was detected in all samples, while the K/Nb molar ratio of K4 Nb6 O17 was found to be much higher than that of CdS-intercalated samples. This may suggest that most of K+ was replaced by H+ , while some K+ still exists in the material during the ion exchange reaction, which is in agreement with literature [21]. The K/Nb molar ratio of K4 Nb6 O17 was 74.97:100, which was slightly higher than its stoichiometric ratio of 4:6, due to the excessive K2 CO3 dosage via the high temperature solid phase reaction. Furthermore, the K/Nb molar ratios of mm-K4 Nb6 O17 /CdS and mn-K4 Nb6 O17 /CdS were much lower than those of nm-K4 Nb6 O17 /CdS and nn-K4 Nb6 O17 /CdS, indicating that microwave irradiation could provide a higher efficiency for H+ exchange reaction than the conventional synthesis. Meanwhile, the Cd/Nb molar ratio in mm-K4 Nb6 O17 /CdS was higher than other as-prepared intercalated samples, indicating that microwave irradiation could also promote the intercalation CdS. Optical absorption of the as-prepared K4 Nb6 O17 , mmK4 Nb6 O17 /CdS, M-CdS/K4 Nb6 O17 (direct mixture of CdS and K4 Nb6 O17 ) and CdS photocatalysts were investigated using a

W. Cui et al. / Applied Catalysis A: General 417–418 (2012) 111–118

115

Table 1 XRF results of the samples prepared via different methods. Molar ratio

mm-K4 Nb6 O17 /CdS

mn-K4 Nb6 O17 /CdS

nm-K4 Nb6 O17 /CdS

nn-K4 Nb6 O17 /CdS

K4 Nb6 O17

K:Nb Cd:Nb

7.91:100 31.94:100

7.71:100 30.92:100

11.35:100 28.64:100

11.35:100 27.01:100

74.97:100 0:100

UV–vis spectrometer. As shown in Fig. 3(A), the K4 Nb6 O17 sample absorbs UV light, and the absorption edge is at 400 nm, corresponding to a bandgap of approximately 3.1 eV, which cannot be present in visible light. The absorption edge of pure CdS was determined to be 550 nm, corresponding to the direct-bandgap of 2.4 eV, which agreed with the literature [22]. Compared to K4 Nb6 O17 , it was apparent that the absorption edge of M-CdS/K4 Nb6 O17 and mm-K4 Nb6 O17 /CdS appeared redshifted into the visible light region. Since K4 Nb6 O17 does not absorb in the visible range, these red-shifts could be attributed to distinct specimens of CdS in the samples. The absorption edge of mm-K4 Nb6 O17 /CdS was found to be at about 530 nm, which was shorter that that of pure CdS and M-CdS/K4 Nb6 O17 , and the blue-shift can be attributed to the small particle size of CdS in mm-K4 Nb6 O17 /CdS [23]. Therefore, K4 Nb6 O17 /CdS photocatalysts could be excited to produce more electron–hole pairs under the same amount of irradiation, which could result in the higher photocatalysitic activity. Fig. 3(B) shows the absorption spectras of CdS-intercalated K4 Nb6 O17 samples prepared via different methods. As shown, all samples exhibited similar diffuse absorption spectra, indicating that there are little differences in the band structures of the samples. Molecular fluorescence spectroscopy is a kind of emission spectrum caused by the electron–hole recombination, which can reflect the migration and capture of photo-induced carriers [24]. The results obtained for the prepared samples are shown in Fig. 4. Pure K4 Nb6 O17 showed one peak at around 375 nm and pure CdS showed one peak at around 500 nm. All composite catalysts exhibited two peaks at 375 nm and 540–560 nm, respectively. As shown in Fig. 4(A), all samples showed blue luminescence with a maximum around 375 nm, which is attributed to the emission of K4 Nb6 O17 according to the literature [25]. Meanwhile, the emission intensity drastically decreased with the intercalation of CdS. This may have indicated that the transition of photo-induced electrons from CB to VB in the bulk of K4 Nb6 O17 was blocked and the recombination of electron–hole pairs was suppressed due to the interaction of CdS and K4 Nb6 O17 . The interaction of CdS can

3.2. Photocatalytic activity The photocatalytic activities of the as-prepared samples were quantified by the rate of hydrogen production under UV light irradiation for 3 h. The activities of hydrogen production over different K4 Nb6 O17 /CdS photocatalysts, as well as pure K4 Nb6 O17

Absorbance intensity (a. u.)

B

Absorbance intensity (a. u.)

A

promote the separation of photo-induced hole–electron pairs [26]. Furthermore, the intensity of the emission peak decreased with in the following order: nn-K4 Nb6 O17 /CdS, nm-K4 Nb6 O17 /CdS, mnK4 Nb6 O17 /CdS and mm-K4 Nb6 O17 /CdS. The could be explained by the different amount of intercalated CdS via different methods. mm-K4 Nb6 O17 /CdS showed the weakest peak at 375 nm, indicating that the highest proportion of intercalated CdS was achieved via microwave irradiation. The strong peak at around 500 nm is assigned to the emission of pure CdS (curve a in Fig. 4(B)), which was consistent with reports in literature [27]. When CdS particles were intercalated into K4 Nb6 O17 , the photocatalysts exhibited emission peaks at 540–560 nm (curves b, c, d and e), which were red-shifted compared to pure CdS and assigned to radiative recombination of charge carriers trapped in a narrower band structure coupled between K4 Nb6 O17 and CdS. The photo-induced electrons in the CB of K4 Nb6 O17 could be transferred from the CB of K4 Nb6 O17 to the VB of CdS and encounter with the photo-induced holes, because the CB energy level of K4 Nb6 O17 is lower than that of CdS and the VB energy level of CdS is higher than that of K4 Nb6 O17 [19,28], which could result in the red-shifted emission. Furthermore, these redshifted peaks were weaker than those of pure CdS, indicating that the recombination of charge carriers was efficiently suppressed due to the electric field which existed at the junction between the two semiconductors. According to Figs. 3 and 4, it could be suggested that a combined band structure was formed between CdS and K4 Nb6 O17 , which could effectively improve the separation of the photogenerated electrons–holes [26]. Additionally, the weakest peak at 540–560 nm was observed in mm-K4 Nb6 O17 /CdS sample, which could be due to the high degree of crystallinity.

d c b

a 300

d

c

b a

400

500 λ/ nm

600

700

300

400

500

600

700

λ / nm

Fig. 3. UV–vis diffuse absorption spectra of photocatalysts. (A) ((a) K4 Nb6 O17 (b) mm-K4 Nb6 O17 /CdS (c) M-CdS/K4 Nb6 O17 (d) CdS); (B) ((a) mm-K4 Nb6 O17 /CdS (b) mnK4 Nb6 O17 /CdS (c) nm-K4 Nb6 O17 /CdS (d) nn-K4 Nb6 O17 /CdS).

116

W. Cui et al. / Applied Catalysis A: General 417–418 (2012) 111–118

A

B

PL Intensity/a.u

b

a

PL Intensity/a.u

a

d

c

e c

b

d

e 250

300

350

440

400

460

480

500

520

540

560

580

600

λ/nm

λ/nm

Fig. 4. Fluorescence spectra of photocatalysts. (A) ((a) K4 Nb6 O17 (b) nn-K4 Nb6 O17 /CdS (c) nm-K4 Nb6 O17 /CdS (d) mn-K4 Nb6 O17 /CdS (e) mm-K4 Nb6 O17 /CdS); (B) ((a) CdS (b) nn-K4 Nb6 O17 /CdS (c) nm-K4 Nb6 O17 /CdS (d) mn-K4 Nb6 O17 /CdS (e) mm-K4 Nb6 O17 /CdS).

and CdS are given in Fig. 5(A). As shown, all photocatalysts exhibited photo-activities under UV light irradiation. The photoactivity for H2 production over the M-CdS/K4 Nb6 O17 for 3 h irradiation was 34.843 mmol/(g cat), which was lower than that of K4 Nb6 O17 . This suggested that CdS shadowed the surface of K4 Nb6 O17 particles in the M-CdS/K4 Nb6 O17 material [29]. Only

13.59 mmol/(g cat) hydrogen was produced over pure CdS, because it was easily photo-corroded in an aqueous solution [30]. Overall, the CdS intercalated K4 Nb6 O17 exhibited higher activities than pure K4 Nb6 O17 , pure CdS and M-CdS/K4 Nb6 O17 . This may suggest that the two kinds of band structures combined between CdS and K4 Nb6 O17 in the CdS intercalated K4 Nb6 O17 photocatalysts was

B

A

6

250

200

H2 production mmol/(g cat)

H2 production mmol/(g cat)

5

150

100

50

4

3

2

1

0

0

a

b

c

d

H2 production / mmol/(g cat)

C

e

f

g

a

b

c

d

e

f

g

6 nd

st

2 run

1 run

rd

3 run

th

th

4 run

5 run

4

2

0 0

3

6 9 Irradiation time / h

12

15

Fig. 5. H2 production of photocatalysts. (A) Under UV irradiation; (B) under visible light irradiation; (C) repeated runs of photocatalysis over recycled sample d under visible light. ((a) K4 Nb6 O17 ; (b) nn-K4 Nb6 O17 /CdS; (c) mn-K4 Nb6 O17 /CdS; (d) mm-K4 Nb6 O17 /CdS; (e) nm-K4 Nb6 O17 /CdS; (f) M-CdS/K4 Nb6 O17 ; (g) CdS).

W. Cui et al. / Applied Catalysis A: General 417–418 (2012) 111–118

A

-2

B e

-1

UV

-2

e-

-1

CdS

visible light

CdS K4Nb6O17

K4Nb6O17

0

2.4eV e-

3.1eV e-

1

V/NHE

V/NHE

117

0

2.4eV e1

+ h

3.1eV

h+

2

+ h

+ h

3

2

3

Fig. 6. Photogenerated electrons–holes transition spectra. (A) Under UV irradiation; (B) under visible light irradiation.

an important factor for the improvement of photocatalytic efficiency. The combined band structure inhibits the recombination of photo-generated electron–hole pairs and greatly enhances the photocatalytic activity. Among the CdS intercalated K4 Nb6 O17 photocatalysts, mm-K4 Nb6 O17 /CdS achieved the highest activity, with 256.95 mmol/(g cat) H2 produced under this condition; followed by mn-K4 Nb6 O17 /CdS (248.76 mmol/(g cat)). nm-K4 Nb6 O17 /CdS and nn-K4 Nb6 O17 /CdS showed lower activities of 187.66 mmol/(g cat) and 168.6 mmol/(g cat), respectively. The photocatalysis results indicated that the preparation method had a great influence on the photocatalytic performance of CdS intercalated K4 Nb6 O17 . Microwave irradiation introduced in the intercalation preparation process, especially in regards to H+ exchange and butylamine intercalation, could improve the photocatalytic activity. This was supported by the fact that, among the samples, the mmK4 Nb6 O17 /CdS exhibited a similar activity with mn-K4 Nb6 O17 /CdS, which was much higher than that of the nm-K4 Nb6 O17 /CdS and nn-K4 Nb6 O17 /CdS samples, in agreement with the XRF results. Fig. 5(B) shows the photocatalytic activities of H2 production over as-prepared photocatalysts under visible light irradiation. The photocatalyst samples under visible light irradiation showed the similar relative activities to those observed under UV light irradiation. The pure K4 Nb6 O17 exhibited no activity for hydrogen evolution under visible light irradiation, due to its wide bandgap of 3.1 eV. The pure CdS and M-CdS/K4 Nb6 O17 showed photo-activities, and about 0.49 mmol/(g cat) and 0.72 mmol/(g cat) hydrogen were produced after 3 h irradiation. The activity of MCdS/K4 Nb6 O17 was higher than that of pure CdS, indicating that the combined band structure in M-CdS/K4 Nb6 O17 could promote the separation of photo-generated electron–hole pairs. When CdS particles were intercalated into the interlayer of K4 Nb6 O17 , the materials showed higher photocatalytic activity for hydrogen evolution than those of pure CdS and M-CdS/K4 Nb6 O17 under the visible light irradiation. This indicated that the intercalation of CdS could improve the photocatalytic activity by efficient separation of the hole–electron pairs which were induced on CdS under visible irradiation. The photocatalytic activities of H2 production over mm-K4 Nb6 O17 /CdS, mn-K4 Nb6 O17 /CdS, nm-K4 Nb6 O17 /CdS and nn-K4 Nb6 O17 /CdS for 3 h irradiation were 5.68 mmol/(g cat), 4.68 mmol/(g cat), 2.97 mmol/(g cat) and 2.34 mmol/(g cat) respectively. The highest photocatalytic activity was achieved over the mm-K4 Nb6 O17 /CdS catalyst, which was prepared using microwave irradiation-assisted intercalation. Stability is an important parameter for photocatalysts. To evaluate the stability and reusability of the photocatalyst, we performed repeated runs for the photocatalytic hydrogen evolution over mmK4 Nb6 O17 /CdS under visible light, recycling the catalyst between runs. As shown in Fig. 5(C), after five cycles of photocatalysis, the

as-prepared catalyst did not exhibit any significant loss of activity, indicating it possessed good stability for repeated photocatalytic reaction. 3.3. Mechanism analysis The photocatalytic mechanism was analyzed from the aspect of the transition of the photogenerated electrons–holes. The valence band and conduction band of CdS were determined by the 3p orbital of the S atom and the 5s orbital of the Cd atom, respectively, and the valence band and conduction band of K4 Nb6 O17 were determined by the 2p orbital of the O atom and the 4d orbital of the Nb atom, respectively [31,32]. In addition, the energy level of the 5s orbital of the Cd atom was higher than the 4d orbital of the Nb atom and the energy level of the 3p orbital of the S atom was higher than the 2p orbital of the O atom [19,28]. According to the relative position of the valence band and conduction band of CdS and K4 Nb6 O17 , the band structure of the CdS intercalated K4 Nb6 O17 was determined, and is depicted in Fig. 6. As shown in Fig. 6(A), CdS and K4 Nb6 O17 both generated electrons and holes under UV irradiation. The photogenerated electrons on the conduction band of CdS could migrate to the conduction band of K4 Nb6 O17 with the higher potential, and the photogenerated holes on the valence band of K4 Nb6 O17 could migrate to the valence band of CdS with the lower potential. Thus, photogenerated electrons aggregated on the conduction band of K4 Nb6 O17 and holes aggregated on the valence band of CdS, which facilitated the separation of the photogenerated electron–hole pairs and acted to improve the overall photocatalytic activity. Under visible light irradiation, as shown in Fig. 6(B), only CdS powders generated electrons and holes. The photogenerated electrons on the conduction band of CdS could migrate to the conduction band of K4 Nb6 O17 with the higher potential. This also promoted the separation of photogenerated electrons–holes and increased the photocatalytic activity. 4. Conclusions CdS intercalated K4 Nb6 O17 photocatalysts were successfully synthesized using a microwave-assisted synthesis. CdS/K4 Nb6 O17 photocatalyst synthesized under microwave irradiation retained a higher crystallinity than the same catalyst synthesized via conventional synthesis. The absorption edge of K4 Nb6 O17 shifted to the visible light region after the intercalation of CdS. Compared to conventional synthesis, the use of microwave irradiation sharply shortened the intercalation time. Further, the CdS/K4 Nb6 O17 sample prepared via microwave irradiation exhibited higher activities for photocatalytic hydrogen production under UV light and

118

W. Cui et al. / Applied Catalysis A: General 417–418 (2012) 111–118

visible light irradiation, and the amounts of hydrogen produced were 265.95 mmol/(g cat) and 5.68 mmol/(g cat) after 3 h, respectively. Acknowledgments This work was supported by the National Natural Science Foundation of China (grant no. 50972037, no. 51172063), the Research Project of Education Department of Hebei Province (no. 2010130). The authors also thank Joanne Gamage and Prof. Z. Zhang, Department of Chemical & Biological Engineering, University of Ottawa, for critical reading of the manuscript. References [1] G.L. Chiarello, D. Ferri, E. Selli, J. Catal. 280 (2011) 168–177. [2] R. Sasikala, A.R. Shirole, V. Sudarsan, V.S. Kamble, C. Sudakar, R. Naik, R. Rao, S.R. Bharadwaj, Appl. Catal. A: Gen. 390 (2010) 245–252. [3] O. Akhavan, R. Azimirad, Appl. Catal. A: Gen. 369 (2009) 77–82. [4] M.T. Niu, F. Huang, L.F. Cui, P. Huang, Y.L. Yu, Y.S. Wang, ACS Nano 4 (2010) 681–688. [5] K. Maeda, M. Eguchi, W.J. Youngblood, T.E. Mallouk, Chem. Mater. 20 (2008) 6770–6778. [6] M.R. Allen, A. Thibert, E.M. Sabio, N.D. Browning, D.S. Larsen, F.E. Osterloh, Chem. Mater. 22 (2010) 1220–1228. [7] L. Liu, W.Q. Cui, Y.H. Liang, F.L. Qiu, Acta Chim. Sinica 68 (2010) 211–216. [8] T. Nakato, H. Edakubo, T. Shimomura, Microporous Mesoporous Mater. 123 (2009) 280–288. [9] J. Yuan, W.G. Tao, W.F. Shanguan, J.W. Shi, Z.Y. Liu, Rare Met. Mater. Eng. 33 (2004) 219–221. [10] S. Tawkaew, Y. Fujishiro, S. Yin, T. Sato, Colloids Surf. A 179 (2001) 139–144.

[11] M.A. Bizeto, V.R.L. Constantino, Mater. Res. Bull. 39 (2004) 1811–1820. [12] U. Unal, Y. Matsumoto, N. Tamoto, M. Koinuma, M. Machida, K. Izawa, J. Solid State Chem. 179 (2006) 33–40. [13] J.J. Wu, X.J. Lv, L.L. Zhang, Y.J. Xia, F.Q. Huang, F.F. Xu, J. Alloys Compd. 496 (2010) 234–240. [14] W.W. Qu, F. Chen, B. Zhao, J.L. Zhang, J. Phys. Chem. Solids 71 (2010) 35–41. [15] X.R. Zhang, X. Zheng, Mater. Lett. 61 (2007) 1478–1482. [16] A. Olaya, S. Moreno, R. Molina, Catal. Commun. 10 (2009) 697–701. [17] L.L. Wang, B. Li, C.X. Chen, L.N. Jia, J. Alloys Compd. 508 (2010) 426–432. [18] T. Nakato, M. Kameyama, Q.M. Wei, J. Haga, Microporous Mesoporous Mater. 110 (2008) 223–231. [19] S. Uchida, Y. Yamamoto, Y. Fujishiro, A. Watanabe, O. Ito, T. Sato, J. Chem. Soc. Faraday Trans. 93 (1997) 3229–3234. [20] K. Domen, Y. Ebina, S. Ikeda, et al., Catal. Today 28 (1996) 167–174. [21] A. Furebe, T. Shiozawa, A. Ishikawa, A. Wada, C. Hirose, K. Domen, Chem. Phys. 285 (2002) 31–37. [22] F.A. Majumder, H.E. Swoboda, K. Kempf, C. Klingshirn, Phys. Rev. B 32 (1985) 2407–2418. [23] J.G. Yu, J. Zhang, M. Jarnoniec, Green Chem. 12 (2010) 1611–1614. [24] H.W. Chen, Y. Ku, Y.L. Kuo, Water Res. 41 (2007) 2067–2078. [25] M.C. Sarahan, E.C. Carroll, M. Allen, D.S. Larsen, N.D. Browing, F.E. Osterloh, J. Solid State Chem. 181 (2008) 1678–1683. [26] Q.Q. Wang, B.Z. Lin, B.H. Xu, L.L. Xiao, Z.J. Chen, X.T. Pian, Microporous Mesoporous Mater. 130 (2010) 344–351. [27] A. Chatterjee, A. Priyam, S.C. Bhattacharya, A. Saha, J. Lumin. 126 (2007) 764–770. [28] G.Q. Guan, T. Kida, K. Kusakabe, K. Kimura, E. Abe, A. Yoshida, Appl. Catal. A: Gen. 295 (2005) 71–78. [29] H.Y. Lin, T.H. Lee, C.Y. Sie, Int. J. Hydrogen Energy 33 (2008) 4055–4063. [30] D.K. Koll, A.H. Taha, D.M. Giolando, Sol. Energy Mater. Sol. Cells 95 (2011) 1716–1719. [31] E.C. Ekuma, L. Franlin, G.L. Zhao, J.T. Wang, D. Bagayoko, Phys. B 406 (2011) 1477–1480. [32] Q.Y. Li, T. Kaka, J.H. Ye, Int. J. Hydrogen Energy 36 (2011) 4716–4723.