Incorporation of Polyoxotungstate Complexes in Silica Spheres and

The resulted silica composites were named as S-0, S-1, S-2, S-3, S-4, and S-5 .... DLS measurement demonstrates the existence of CTAB/SEP-1 aggregates...
2 downloads 0 Views 5MB Size
Download PDF
pubs.acs.org/Langmuir © 2010 American Chemical Society

Incorporation of Polyoxotungstate Complexes in Silica Spheres and in Situ Formation of Tungsten Trioxide Nanoparticles Yuanyuan Zhao,† Haimei Fan,‡ Wen Li,† Lihua Bi,† Dejun Wang,‡ and Lixin Wu*,† †

State Key Laboratory of Supramolecular Structure and Materials, ‡Alan G. MacDiarmid Institute College of Chemistry, and Jilin University, Changchun 130012, P. R. China Received June 18, 2010. Revised Manuscript Received July 29, 2010

In this paper, we demonstrated a new convenient route for in situ fabrication of well separated small sized WO3 nanoparticles in silica spheres, through a predeposition of surfactant encapsulated polyoxotungates as tungsten source, and followed by a calcination process. In a typical procedure, selected polyoxotungates with different charges were enwrapped with dioctadecyldimethylammonium cations through electrostatic interaction. Elemental analysis, thermogravimetric analysis, and spectral characterization confirmed the formation of prepared complexes with the anticipated chemical structure. The complexes were then phase-transferred into aqueous solution that predissolved surfactant cetyltrimethylammonium bromide, and finally incorporated into silica spheres through a joint sol-gel reaction with tetraethyl orthosilicate in a well dispersed state under the protection of organic layer for polyoxotungates from the alkaline reaction condition. Transmission electron microscopic images illustrated the well dispersed WO3 nanoparticles in the size range of ca. 2.2 nm in the silica spheres after the calcination at 465 °C. The sizes of both the silica spheres and WO3 nanoparticles could be adjusted independently through changing the doping content to a large extent. Meanwhile, the doped polyoxotungate complexes acted as the template for the mesoporous structure in silica spheres after the calcination. Along with the increase of doping content and surfactant, the mesopore size changed little (2.0-2.9 nm), but the specific surface areas increased quite a lot. Importantly, the WO3-nanoparticle-doped silica spheres displayed an interesting photovoltaic property, which is favorable for the funtionalization of these nanomaterials.

Introduction Polyoxometalates (POMs), as a family of nanosized metal-oxygen molecular clusters with diverse composition and topology, have received considerable attention recently due to their potential applications in catalysis, medicine, optics, and molecular devices.1 To integrate the functional features with the desired dispersivity, porosity, size controllability, and biocompatibility, POMs were tried to incorporated into various carriers such as activated carbon, polymers, molecular sieves, and nanoparticles.2 However, among those carriers, silica spheres doped with POMs in a stable chemical microenvironment without leakage, aggregation, and decomposition are rarely reported,3 because of the pH sensitivity under one pot sol-gel reaction conditions.3a Though the impregnation technique is widely used, it is still difficult to avoid the aggregation and leakage of POMs from the matrix.3b-e To overcome these disadvantages, we developed an effective approach for preparation of POM-doped silica nanospheres by wrapping POMs with hydroxyl group terminaled organic quarternary ammonium cations through electrostatic interaction. The obtained surfactant encapsulated *To whom correspondence should be addressed. E-mail: [email protected]. (1) Special issue on “Polyoxometalates”. Chem. Rev. 1998, 98, 1-390. (2) (a) Inumaru, K.; Ishihara, T.; Kamiya, Y.; Okuhara, T.; Yamanaka, S. Angew. Chem., Int. Ed. 2007, 46, 7625–7628. (b) Qi, W.; Wang, Y. Z.; Li, W.; Wu, L. X. Chem.-Eur. J. 2009, 16, 1069–1078. (c) Wu, Q. Y.; Lin, H. H.; Meng, G. Y. J. Solid State Chem. 1999, 148, 419–424. (d) Kasai, J.; Nakagawa, Y.; Uchida, S.; Yamaguchi, K.; Mizuno, N. Chem.-Eur. J 2006, 12, 4176–4184. (e) Qi, W.; Li, H. L.; Wu, L. X. Adv. Mater. 2007, 19, 1983–1987. (f) Li, H. L; Li, P.; Yang, Y.; Qi, W.; Sun, H.; Wu, L. X. Macromol. Rapid Commun. 2007, 28, 431–436. (g) Zhao, Y. Y.; Qi, W.; Li, W.; Wu, L. X. Langmuir 2010, 26, 4437–4442. (3) (a) Green, M.; Harries, J.; Wakefield, G.; Taylor, R. J. Am. Chem. Soc. 2005, 127, 12812–12813. (b) Kumar, G. D. K.; Baskaran, S. J. Org. Chem. 2005, 70, 4520– 4523. (c) Zhang, F. M.; Wang, J.; Yuan, C. S.; Ren, X. Q. Catal. Lett. 2005, 102, 171– 174. (d) Said, A. A.; Abd El-Wahab, M. M. M.; Alian, A. M. J. Chem. Technol. Biotechnol. 2007, 82, 513–523. (e) Haber, J.; Pamin, K.; Matachowski, L.; Mucha, D. Appl. Catal. A-Gen. 2003, 256, 141–152.

14894 DOI: 10.1021/la102491a

POMs (SEPs) possess an organic shell which can effectively decreases the destruction of POMs from outside solvents and pH, and the grafted hydroxyl groups on the outside surface of the shell make the SEPs covalently connect to the silica matrix in a well dispersed state.2g Thus, it is of interest to exploit new strategy to functionalize the silica spheres. Tungsten trioxide (WO3) and tungsten oxide based compounds have been proven to be an important type of inorganic material over the past several years, because of their polytyped morphologic sphere structures, intriguing physical and chemical properties, and diverse applications. Compared to the bulk materials, nanostructured WO3 shows superior features such as enhanced electro-4 and photochroism,5 catalysis and reaction,6 and bandgap tuning.7 Various newly developed methods, such as sputtering,8 sol-gel,9 spray pyrolysis,10 depositions,7c,11 thermal oxidation of tungsten powder,12 self-assembly,4b,c,13 and template (4) (a) Lee, S. H.; Deshpande, R.; Parilla, P. A.; Jones, K. M.; To, B.; Mahan, A. H.; Dillon, A. C. Adv. Mater. 2006, 18, 763–766. (b) Santato, C.; Ulmann, M.; Augustynski, J. Adv. Mater. 2001, 13, 511–514. (c) Brezesinski, T.; Rohlfing, D. F.; Sallard, S.; Antonietti, M.; Smarsly, B. M. Small 2006, 2, 1203–1211. (5) Monllor-Satoca, D.; Borja, L.; Rodes, A.; Gomez, R.; Salvador, P. ChemPhysChem 2006, 7, 2540–2551. (6) (a) Koo, D. H.; Kim, M.; Chang, S. Org. Lett. 2005, 7, 5015–5018. (b) Suzuki, K.; Watanabe, T.; Murahashi, S. I. Angew. Chem., Int. Ed. 2008, 47, 2079–2081. (c) Hu, J. C.; Wang, Y. D.; Chen, L. F.; Richards, R.; Yang, W. M.; Liu, Z. C.; Xu, W. Microporous Mesoporous Mater. 2006, 93, 158–163. (7) (a) Georg, A.; Graf, W.; Wittwer, V. Vacuum 2008, 82, 730–735. (b) Kowalski, D.; Aoki, Y.; Habazaki, H. Angew. Chem., Int. Ed. 2009, 48, 7582–7585. (c) Qureshi, U.; Blackman, C.; Hyett, G.; Parkin, I. P. Eur. J. Inorg. Chem. 2007, 1415–1421. (d) Morales, W.; Cason, M.; Aina, O.; de Tacconi, N. R.; Rajeshwar, K. J. Am. Chem. Soc. 2008, 130, 6318–6319. (e) Shengelaya, A.; Reich, S.; Tsabba, Y.; M€uller, K. A. Eur. Phys. J. B 1999, 12, 13–15. (8) Bellac, D. L.; Azens, A.; Granqvist, C. G. Appl. Phys. Lett. 1995, 66, 1715– 1716. (9) (a) Santato, C.; Odziemkowski, M.; Ulmann, M.; Augustynski, J. J. Am. Chem. Soc. 2001, 123, 10639–10649. (b) Wang, Z. C.; Chumanov, G. Adv. Mater. 2003, 15, 1285–1289. (c) Srivastava, A. K.; Agnihotry, S. A.; Deepa, M. Thin Sol. Films 2006, 515, 1419–1423. (d) Lee, K.; Seo, W. S.; Park, J. T. J. Am. Chem. Soc. 2003, 125, 3408–3409.

Published on Web 08/18/2010

Langmuir 2010, 26(18), 14894–14900

Zhao et al. Scheme 1. Schematic Drawings of Overall Procedures of SEPs through Electrostatic Interaction of POMs with DODA, Their Transferring Process into Aqueous Solution through Hydrophobic Interaction with CTAB, Followed Incorporation into Silica Spheres Through Cohydrolization with TEOS, and Finally in Situ Formation of WO3 Nanoparticles

Article

Through a brief calcination, the covered organic layer on POMs pyrolyzes completely and the corresponding SEPs in situ transform into WO3 nanoparticles in hybrid silica particles. Although the preparation of WO3 nanoparticles with POMs as the source has been described elsewhere, the present strategy reveals visible advantages: small sized WO3 nanoparticles without surface modification can be conveniently constructed in situ in silica spheres; the prepared WO3 nanoparticles are well separated in a narrow distribution; and the as-prepared WO3 nanoparticles are located in confined cavities of porous structured silica spheres, which can be applied for catalysis, gas sensors, and other functional devices. Interestingly, such dispersed WO3 nanoparticles in silica spheres exhibit quantum size effects and enhanced photovoltaic properties.

Experimental Section

formation14 from various tungsten precursors, have been applied for the preparation of nanostructured WO3. However, the morphology, crystalline state, and especially the size control are critical factors for the performance of the functional features of WO3, and novel approaches are still desired, especially the method for the preparation of small sized nanoparticles in various carriers. Due to both rich tungsten contents and nanoscaled uniform clusters, some POMs behave as competitive candidates for the precursor of WO3 nanoparticles. As an exclusive example, phosphotungstic acid was employed for nanostructured WO3 through a hard template method.14a,15 Meanwhile, the combined merit of polyoxotungstates serving as the precursors of WO3 nanoparticles is the composition tuning through applying substituted POMs.7 In this regard, avoiding the aggregation of yielded nanoparticles and employing POMs both as the source of tungsten and the template of nanoparticles become current challenges. Considering the well dispersion of SEPs in silica matrices and the aggregation block of the organic layer covered on the surface of polyoxotungstate clusters, herein we demonstrate a new approach for the fabrication of small sized WO3 nanoparticles in silica spheres, as schematically presented in Scheme 1. In a typical process, we cover the POMs with organic quaternary ammoniums with double long alkyl chains first, and then a common single alkyl chain cationic surfactant is used to transfer the SEP complexes into an alkaline sol-gel system. Finally, the co-sol-gel process leads to uniformly SEP-doped SiO2 objects. (10) (a) Enesca, A.; Duta, A. Phys. Stat. Solidi (C) 2008, 5, 3499–3502. (b) Solis, J. L.; Rodriguez, J.; Estrada, W. Phys. Stat. Solidi (A) 2004, 201, 2370–2374. (11) Fang, G. J.; Liu, Z. L.; Sun, G. C.; Yao, K. L. Phys. Stat. Solidi (A) 2001, 184, 129–137. (12) (a) Solis, J. L.; Hoel, A.; Kish, L. B.; Granqvist, C. G.; Saukko, S.; Lantto, V. J. Am. Ceram. Soc. 2001, 84, 1504–1508. (b) Su, C. Y; Lin, C. K.; Yang, T. K.; Lin, H. C.; Pan, C. T. Int. J. Refract. Met. Hard Mater. 2008, 26, 423–428. (c) Shibuya, M.; Miyauchi, M. Adv. Mater. 2009, 21, 1373–1376. (13) (a) Chen, L.; Xu, J.; Fleming, P.; Holmes, J. D.; Morris, M. A. J. Phys. Chem. C 2008, 112, 14286–14291. (b) Guo, Y. F.; Quan, X.; Lu, N.; Zhao, H. M.; Chen, S. Environ. Sci. Technol. 2007, 41, 4422–4427. (c) Wang, Z. X.; Zhou, S. X.; Wu, L. M. Adv. Funct. Mater. 2007, 17, 1790–1794. (14) (a) Rossinyol, E.; Prim, A.; Pellicer, E.; Arbiol, J.; Hernandez-Ramı´ rez, F.; Peiro, F.; Cornet, A.; Morante, J. R.; Solovyov, L. A.; Tian, B. Z.; Bo, T.; Zhao, D. Y. Adv. Funct. Mater. 2007, 17, 1801–1806. (b) Zhao, X. F.; Cheung, T. L. Y.; Zhang, X. T.; Ng, D. H. L.; Yu, J. J. Am. Ceram. Soc. 2006, 89, 2960–2963. (c) Ganesan, R.; Gedanken, A. Nanotechnology 2008, 19, 025702. (d) Oaki, Y.; Imai, H. Adv. Mater. 2006, 18, 1807–1811. (15) Zhu, K. K.; He, H. Y.; Xie, S. H.; Zhang, X.; Zhou, W. Z.; Jin, S.; Yue, B. Chem. Phys. Lett. 2003, 377, 317–321.

Langmuir 2010, 26(18), 14894–14900

Chemicals. Analytical grade cetyltrimethylammonium bromide (CTAB), tetraethyl orthosilicate (TEOS), ammonia (NH3 3 H2O, 28%), and hydrofluoric acid (HF, 55%), as well as ethanol in excellent purity, were purchased from Beijing Chemical Reagents Company. Analytical grade H3PW12O40 3 xH2O (POM-4) was from Sinopharm Chemical Reagent Co., Ltd. Dioctadecyldimethylammonium bromide (DODA 3 Br) was from SigamaAldrich. All of the chemicals were used without any further purification except that TEOS was redistilled under reduce pressure just before use. Ultra pure water (18.2 MΩ) was prepared from a Milli-Q Water System (Millipore) and was used throughout the experiments. Preparation of SEPs. Polyanion clusters, Na9EuW10O36 (POM-1), K12EuP5W30O110 (POM-2), and K12.5Na1.5NaP5W30O110 (POM-3), were synthesized according to the published procedures.16 Typically, we initially dissolved POMs in aqueous solution under a suitable pH condition and DODA 3 Br in chloroform separately, and the two solutions holding a charge ratio of POM to DODA in ca. 1-1.1:1 were then mixed together. After a suitable stirring (2-3 h) at room temperature, the organic phase was separated and washed with water several times and dried over dry Na2SO4, giving the SEP products.17 For SEP-1, nine counterions of POM-1 were fully replaced with DODA, yielding the product in a chemical formula (DODA)9EuW10O36 3 9H2O (7687.9). Anal. Calcd: C, 53.43; H, 9.68; N, 1.64. Found: C, 52.98; H, 9.76; N, 1.63. Similarly, for SEP-2 prepared from POM-2, a chemical formula of (DODA)12EuP5W30O110 (14194.5) was anticipated. Anal. Calcd: C, 38.58; H, 6.82; N, 1.18. Found: C, 38.98; H, 6.47; N, 0.96. For SEP-3, started from POM-3, the expected chemical formula is (DODA)14NaP5W30O110 (15167.7). Anal. Calcd: C, 42.13; H, 7.44; N, 1.29. Found: C, 42.66; H, 7.03; N, 0.93. As a comparison, SEP-4 was prepared from POM-4 in a chemical formula of (DODA)3PW12O40 3 5H2O (4620.3). Anal. Calcd: C, 29.64; H, 5.45; N, 0. 91. Found: C, 29.35; H, 5.07; N, 0.66). Preparation of SEP- and WO3-Doped Silica Spheres. All the SEP-doped silica spheres were prepared through the similar route. As an example, to transfer water insoluble SEP-1 complex into aqueous solution, it was initially dissolved in a bit of chloroform and then added dropwise into CTAB aqueous solution with a 1:1 molar ratio to DODA contained in SEP-1. The solution was stirred at 60-80 °C for ca. 30 min to exclude the residual organic solvent, and thus SEP-1 was fully transferred into the condensed aqueous phase at a selected concentration for the sol-gel reaction with TEOS. A series of CTAB transferred SEP-1 (CTAB/SEP-1) with different concentrations was added into the TEOS contained alkaline sol-gel system. After a suitable time of stirring, SiO2 composite particles (SEP-1/SiO2) with loading amount in 0, 7, 13, (16) (a) Creaser, I.; Heckel, M. C.; Neitz, R. J.; Pope, M. T. Inorg. Chem. 1993, 32, 1573–1578. (b) Sugeta, M.; Yamase, T. Bull. Chem. Soc. Jpn. 1993, 66, 444–449. (17) (a) Kurth, D. G; Lehmann, P.; Volkmer, D.; C€olfen, H.; Koop, M. J.; M€uller, A.; Du Chesne, A. Chem.-Eur. J. 2000, 6, 385–393. (b) Bu, W. F.; Fan, H. L.; Wu, L. X.; Hou, X. L.; Hu, C. W.; Zhang, G.; Zhang, X. Langmuir 2002, 18, 6398–6403.

DOI: 10.1021/la102491a

14895

Article

Zhao et al.

25, 38, and 55 wt %, keeping the molar ratio of SEP-1 to TEOS at ca. 0, 1:1720, 1:860, 1:430, 1:215, and 1:108, were thus obtained. The resulted silica composites were named as S-0, S-1, S-2, S-3, S-4, and S-5 in turn. Another sample, S0 -2, was prepared under the molar ratio of CTAB to DODA in SEP-1 at 2:1 with 13 wt % doping concentration for comparison. The corresponding sintered products (W-SEP-1) of S-series samples at 465 °C for 7 h, were denoted as W-0, W-1, W-2, W-3, W-4, W-5, and W0 -2 orderly. For the doped silica spheres of other SEP complexes, SEP-2/SiO2, SEP-3/SiO2, and SEP-4/SiO2, the corresponding sintered products W-SEP-2/SiO2, W-SEP-3/SiO2, and W-SEP4/SiO2 were obtained, accompanied by the addition of same molar ratio of CTAB at 1:1 with DODA contained in these SEPs, respectively, and examined by IR spectra. The molar ratio of SEP2 and -3 to TEOS at 1:860, was employed in the preparation of SEP-2/SiO2 and SEP-3/SiO2, and the molar ratio of SEP-4 to TEOS keeping at 1:176 was adopted in obtaining SEP-4/SiO2. Accordingly, the calcinations for them were performed around 600 °C for 7 h. Measurements. 1H NMR spectra were recorded on a Bruker UltraShieldTM 500 MHz spectrometer. The elmental analysis (EA) result was performed on a Flash EA1112 from ThermoQuest Italia SPA. The thermalgravimetric analysis (TGA) was carried out on a Perkin-Elmer 7 series thermal analysis system. Fourier transform infrared (IR) spectra were recorded on a Bruker IFS66 V FT-IR spectrometer equipped with a DGTS detector (32 scans) at a resolution of 4 cm-1. UV-visble diffuse reflectance spectra (UV-DRS) were carried out on a Shimadzu UV-3100 spectrophotometer. The inductively coupled plasmaoptical emission spectrometer (ICP-OES) used to determine the concentration of a wide range of elements in solutions was Thermo Scientific iCAP ICP-OES 6000 Series. X-ray photoelectron spectroscopic (XPS) measurements were performed on an ES-CALAB Mark (VG, U.K.) photoelectron spectrometer using a monochromatic Al KR X-ray source. Scanning electron microscopic (SEM) images were acquired on a JEOL FESE-M 6700F electron microscope. Transmission electron microscopic (TEM) measurements were finished on a JEOL-2010 electron microscope operating at 200 KV. Raman spectra were recorded on a laser confocal Raman microscope (Renishaw 1000 model) equipped with a CCD detector and a holographic notch filter with excitation laser of 514.5 nm. ζ potential and size measurements were recorded on a Nano ZS (Red badge) ZEN3600 from Malvern. X-ray diffraction (XRD) data were collected on a Rigaku X-ray diffractometer (D/max rA, using Cu KR radiation at a wavelength of 1.542 A˚). The nitrogen adsorption and desorption isotherms at liquid-nitrogen temperature were performed on a Micromeritics ASAP 2020 M system. The calcined samples were degassed for 10 h at 300 °C (samples S-1 to S-5 were degassed for 8 h at 180 °C) before measurements. The surface photovoltage (SPV) measurements were performed on the detection system composed of a source of monochromatic light, a lock-in amplifier (SR830-DSP) with a light chopper (SR540), a photovoltaic cell, and a computer. A 500-W xenon lamp (CHFXQ500W, Global xenon lamp power) and a double-prism monochromator (Hilger and Watts, D300) provided monochromatic light. A low chopping frequency of ∼23 Hz was used. A sandwich-like structure in which the samples were directly located between two blank ITO electrodes was employed.18

Results and Discussion Preparation and Structure Characterization of SEPs. As the four employed POMs possess different negative charges, metal substitution and structural type, and stabilized pH range, we believe that they covered general structural characteristics and properties of POMs. From EA data of all SEPs, it is confirmed (18) Wang, P.; Xie, T. F.; Li, H. Y.; Peng, L.; Zhang, Y.; Wu, T. S.; Pang, S.; Zhao, Y. F.; Wang, D. J. Chem.;Eur. J. 2009, 15, 4366–4372.

14896 DOI: 10.1021/la102491a

Figure 1. (a) TGA of SEP-1 and SEP-1/SiO2 (S-1) and (b) IR spectra of POM-1, SEP-1, and CTAB/SEP-1 in KBr.

that the counterions of POM-1, -2, -3, and -4 are fully replaced by DODA cations, generating anticipated SEP-1 complex in chemical structure (DODA)9EuW10O36 3 9H2O, SEP-2 in (DODA)12EuP5W30O110, SEP-3 in (DODA)14NaP5W30O110, and SEP-4 in (DODA)3PW12O40 3 5H2O. For SEP-3, because the central sodium ion is encapsulated by a cyclic arrangement of five -PW6O22units, similar to the case of europium in POM-2, it becomes the part of the POM anion and can not be substituted by DODA. TGA of SEP-1 (Figure 1a) shows the weight loss of 1.6% in the range from room temperature to 200 °C, matching 2.1% crystalline water estimated from the chemical formula. The measured weight loss of 32.4% between 200 and 840 °C, derived from thermal decomposition, just matches the content of the organic component (32.5%) in the complex by assuming the inorganic residual to be WO3 and Eu2O3. Considering the possible errors, the measured weight losses of 0.4% for SEP-2, 0.2% for SEP-3, and 1.5% for SEP-4 in the range from room temperature to 200 °C are also consistent with the corresponding calculated crystalline waters, 0, 0, and 2.0%, based on the anticipated molecular formulas. Similarly, supposing the calcinated inorganic residues to be WO3 and Eu2O3 for SEP-2, Na2O and WO3 for SEP-3, and WO3 for SEP-4, the calculated residual weight from each chemical formula should be 50.2, 46.1, and 60.2%, which are in good accordance with the experimental values of 50.7, 45.4, and 59.7% at ca. 600 °C. All TGA analysis and EA data were shown in Supporting Information (Figure S1 and Table S1). The IR spectra support the expected electrostatic encapsulation and the anticipated chemical structures of as-prepared complexes. The main typical vibrations of POM-1 appearing at 945 and 895 cm-1, assigned to vas (W-Od) and vas (W-Ob-W), and 843, 778, and 703 cm-1, attributed to vas (W-Oc-W) stretching modes, respectively, were observed clearly in the spectra of the corresponding SEP-1 complex (Figure 1b), implying that the chemical structure of POM-1 was kept in SEP-1 as well. A few wavenumber changes of those vibration bands are attributed to the substitution of surface cations of POM-1, as found in the reported results.2g Meanwhile, the vibrations of DODA at 2920, 2850, and 1468 cm-1, derived from antisymmetric and symmetric stretchings, as well as the distortion mode of methylene groups, were found in the complex, further indicative of the successful encapsulation. The other complexes, SEP-2 and SEP-3, present similar vibrations at those wavenumbers of POM-2 and POM-3, as typically shown in the comparison of IR spectra between SEP-4 and POM-4 (Table S2 and Figure S2). Phase-Transfer of SEPs and Their Aggregation in CTAB Aqueous Solution. To increase the solubility in the ethanol aqueous solution and the dispersion ability in the silica matrix, we tried to modify the end group of DODA with hydroxyl for getting complexes with -OH periphery. However, the amount of SEPs embedded in silica spheres can not reach a much higher level conveniently due to the lower solubility of the complexes. Because Langmuir 2010, 26(18), 14894–14900

Zhao et al.

Figure 2. (a) DLS curve and inserted SEM image of CTAB/SEP-1 aggregates in water and (b) XPS spectra of S-1 before and after etching by 120 s of Ar ion and 1% of HF aqueous solution.

of the enhancement of surfactant for the solubility of hydrophobic objects in aqueous solution, the addition of CTAB effectively promotes the transferring of SEPs into water system, as reported in the case of some transportation of hydrophobic nanoparticles into water.19 Under the employed preparation condition, the CTAB concentration is close to its critical micelle concentration (cmc, ca. 1.0 mmol dm-3) in water.20 The cmc value should increase quite much comparing to the employed CTAB concentration when adding ethanol into the aqueous solution during the sol-gel reaction. Thus, we believe that CTAB could not aggregate itself into micelles but inserts its hydrophobic tail into the organic layer of SEPs, in which the hydrophilicity of cationic head ensures the transference and well dispersion of CTAB incorporated complexes in water. Importantly, the high content of SEPs to be incorporated into silica particles can be well realized through the present approach (Scheme 1). DLS measurement demonstrates the existence of CTAB/SEP-1 aggregates with size around 94 nm in aqueous solution, which is further confirmed by consistent SEM observation, as shown in Figure 2a. Because the measured size of the aggregates is larger than the normal CTAB micelles, it is definite that the observed aggregates are derived from the CTAB incorporated SEP-1 complexes. The ζ potential measurement reveals a positive surface (þ49.7 mV, Figure S3) for the aggregates, supporting that CTAB inserts into the SEP complex with its hydrophobic end and leaves a quaternary ammonium ion on the periphery of the SEP-1 aggregations, which ensures the phase-transfer of SEP-1 into water. Normally, the aggregations are favorable to be embedded in silica spheres because the objects, served as potential seeds, are easily covered by SiO2 during the hydrolysis and polymerization of TEOS. Preparation, Structure, and Morphology of SEP-Doped Silica Particles. On the other hand, the cationic surface of CTAB inserted SEP complexes promotes their incorporation into silica spheres, as well. Also, due to the obstruction of organic layer covered on POMs, the complexes can be well dispersed in isolated state, and the amount to be doped in the silica spheres can reach a high level without an adverse aggregation. Actually, the formation of SEP-1/SiO2 with SEP-1 dispersing uniformly is quite understandable based on the cooperation mechanism.21 The negative oligomers from hydrolyzed TEOS interact with positive cations of CTAB around SEP-1, aggregate and cross-link together. The hydrophobic interaction between SEP-1 and CTAB (19) Kim, J.; Lee, J. E.; Lee, J.; Yu, J. H.; Kim, B. C.; An, K.; Hwang, Y.; Shin, C. H.; Park, J. G.; Kim, J.; Hyeon, T. J. Am. Chem. Soc. 2006, 128, 688–689. (20) (a) Howe, L.; Zhang, J. Z. Photochem. Photobiol. 1998, 67, 90–96. (b) Yao, J.; Tjandra, W.; Chen, Y. Z.; Tam, K. C.; Ma, J.; Soh, B. J. Mater. Chem. 2003, 13, 3053– 3057. (21) Monnier, A.; Sch€uth, F.; Huo, Q.; Kumar, D.; Margolese, D.; Maxwell, R. S.; Stucky, G. D.; Krishnamurty, M.; Petroff, P.; Firouzi, A.; Janicke, M.; Chmelka, B. F. Science 1993, 261, 1299–1303.

Langmuir 2010, 26(18), 14894–14900

Article

tail, the electrostatic interaction between ionic head of CTAB and the hydrolyzed TEOS, as well as the condensability of hydrolyzed TEOS contributes to the final formation of SEP-1/SiO2 with POM-1 confined in a hydrophobic nanoenvironment of alkyl chains, immobilized and uniformly dispersed in Silica matrix. Except Si, O, and trace residual C, the XPS analysis of S-1 shows signals (N element) of the complexes adsorbing on outside surface of silica spheres after a 120 s of Argon ion etching. In contrast to that, the XPS spectrum confirms the existence of N, W, and Eu after etching with 1% HF solution (Figure 2b). The ζ potential measurement of S-1 exhibits a negatively charged surface (-49.2 mV, Figure S3), completely different from the CTAB/SEP-1 aggregates but consistent with that (-46.3 mV) of pure silica particles in water, further indicative of full encapsulation of CTAB/SEP-1 in silica spheres. Under the selected SEP-1 doping concentration, the silica spheres were prepared and the SEM observation was shown in Figure 3a and Figure S4. As an example, the image demonstrates the separated silica spheres with smooth surface and narrow dispersion with a size of ca. 175 nm for S-1 evidently (Figure S5a). From the TEM image of S-1 (Figure 3b), it can be seen that the SEP-1 complex is well distributed in silica spheres. The detailed experimental results confirm that the present method is applicable for all of the complexes under the given doping concentrations. Importantly, no remarkable aggregation of SEPs was found with the increase in doping concentrations of SEP-1 from S-1 to S-5. In addition, both TGA of solid S-1 and ICP analysis of metal ions in the residual solution of S-1 reaction system imply that almost all CTAB/SEP-1s were transferred into hybrid silica spheres during the sol-gel reaction. In the IR spectra of the SEP-1/SiO2 (Figure 4a), the absorption bands assigned to methylene vibrations of alkyl chains were also found at 2920 and 2850 cm-1 clearly, revealing the existence of CTAB covered SEP-1 complex in the silica spheres. The increase of these bands in intensity indicates the increase in the CTAB/ SEP-1 doping concentration from S-1 to S-5. The strong broad absorption band near 1100 cm-1 is obviously derived from the vibrations of Si-O-Si bond through comparing it with the spectrum of blank silica spheres. Also, accompanied by overlapping of Si-O vibrations at 1000-750 cm-1, vas (W-Od) and vas (W-Ob-W) vibrations assigned to POM-1 at the same region can also be recognized evidently. More obviously, the intensities of those vibrations assigned to SEP-1 complex increase with the doping content in the reaction systems from S-1 to S-5 in turn. For other SEP-doped silica spheres, similar results were observed while the alkyl chains get more disordered than each corresponding SEP complex. Hydrophobic Layer Protection of POMs against Chemical Environment. Generally, most naked POMs require a strict pH region due to the cluster stability in solution. For example, POM-1 is usually prepared at pH 7-7.5, and it is unstable outside pH 5.5-8.5,16b,22 which make it difficult for it to be incorporated into silica spheres because of the slightly high pH (8-11) needed in sol-gel reaction. The organic encapsulation of POMs avoids the disadvantage effectively and makes feasible their incorporation into silica spheres. To identify this insight, the chemical structure of POM-1 in SEP-1/SiO2 particles was further evaluated through examining the luminescence of the Eu ion. In comparison to the reported stability of less than 48 h for polyamino acid transferred POM-1,3a S-1 spheres remain luminescent in water longer than 7 months. Moreover, comparing with naked POM-1, (22) Kim, H. S.; Hoa, D. T. M.; Lee, B. J.; Park, D. H.; Kwon, Y. S. Curr. Appl. Phys. 2006, 6, 601–604.

DOI: 10.1021/la102491a

14897

Article

Zhao et al.

Figure 3. (a) SEM image of S-1, TEM images of (b) S-1, (c) W-1, (d) W-2, (e) W-3, (f) W-5, and (g) WO3 nanoparticles through etching W-5 with HF solution, and (h) nanosructured WO3 of W-SEP-1.

Figure 4. IR spectra of (a) CTAB/SEP-1, S-0 and SEP-1/SiO2 with different loading content of SEP-1 and (b) W-SEP-1, W-0, and W-SEP-1/SiO2 with different loading content of SEP-1 in KBr.

Figure 5. Fluorescence spectra of (a) S-1 aqueous solution with similar concentration at pH 1, 5.8 (pure water), and 11.7, excited at 275 nm and (b) POM-1 and CTAB/SEP-1 aqueous solution with the same concentration at pH 1 and 10, excited at 275 nm.

it still maintains considerable intensity even after 36 h of soaking in pH 1 and 11.7 aqueous solutions (Figure 5), indicating the evident shielding performance of the hydrophobic microenvironment for POMs. Therefore, the hydrophobic encapsulation indeed protects POM clusters from the acid and base decompositions. In accordance with SEP-1/SiO2, other SEP complexes were incorporated into silica spheres through the same approach and the structural characterizations demonstrated the feasibility of the present strategy and the formation of SEP-2/SiO2, SEP-3/SiO2, and SEP-4/SiO2 spheres possessing similar size and existing state. In Situ Formation of WO3 Nanoparticles inside Silica Spheres. To carry out the transformation, the SEP-doped silica spheres were allowed to undergo a calcination process. The temperature was chosen according to the TGA of SEP-1. Since POMs are mainly constructed from corner- and edge-sharing WO6 octahedra, it is quite understandable that the WO3 should be mostly the product when decomposition happened at high enough 14898 DOI: 10.1021/la102491a

Figure 6. (a) Raman spectra of W-0, sintered SEP-1 (W-SEP-1), sintered SEP-1 in silica spheres (W-5), and the sample W-5 experienced washing by HCl solution with pH 2 (W-5 in Eu removed state) and (b) XRD patterns of W-SEP-1, W-0 and W-SEP-1/SiO2 spheres with different loading content of SEP-1.

temperature. And, as anticipated, the metal oxides derived from in situ transformation of SEP complexes in silica shperes were obtained during heating up to 465 °C under an O2 atmosphere, where organic components fully decomposed. (Table S3). The IR spectrum (Figure 4b) of W-SEP-1 reveals the exclusion of the organic component and exhibits very similar vibration bands (600-900 cm-1) as that of reported monoclinic WO3.23 Due to the weak absorbance and overlapped vibrations with silica, W-O vibrations derived from WO3 were not observed distinctly. Fortunately, the characteristic vibration modes at 805 and 718 cm-1 assigning to W-O stretching mode, 322 and 272 cm-1 attributing to O-W-O bending modes, together with the weak bands around 437 and 447 cm-1 indicating characteristic bands of crystalline WO3 (Figure 6a), are clearly observed in Raman spectra of all sintered SEP complexes and SEP-doped silica spheres (Figure S6). All these Raman bands indicate the formation of monoclinic tungsten oxide.9a,23 The broad and weak Raman band centered at 969 cm-1 possibly corresponds to the EudO stretching because it did not emerge in the spectra of sintered products from SEP-4 without europium ion and disappeared when we washed the sintered W-5 with HCl. The formation of europium oxide can be further confirmed by fluorescent spectra of sintered samples, which are obviously different from those SEP complexes. Therefore, the formed WO3 should be doped by Eu3þ during the sintering process of europium containing SEP-1/SiO2 spheres. To further verify the final product of SEP-1 after the calcination process, XRD was employed for the characterization as well. In excellent agreement with the diffraction patterns of (23) Daniel, M. F.; Desbat, B.; Lassegues, J. C. J. Solid State Chem. 1987, 67, 235–247.

Langmuir 2010, 26(18), 14894–14900

Zhao et al.

Figure 7. N2 absorption curves, BET surface area, and pore distribution of (a) W-1, (b) W-3, (c) W-SEP-2/SiO2, and (d) W-SEP-3/SiO2.

monoclinic WO3 (JCPDS PDF cards 43-1035; Figure 6b), identical peaks clearly appear in the diffraction patterns of sintered SEP-1, and the intensity for those found diffractions increases with loading of SEP-1 in the silica spheres from W-1 to W-5. Considering the good dispersion of SEP complexes in the silica matrix, it is not surprising that the decomposition product, WO3 nanoparticles, locate inside the cavities that disperse uniformly in silica particles. The sintered silica spheres of S-1 at 465 °C maintain the initial spherical morphology except the surface becoming rough somewhat with some uniform shallow domains dispersing randomly inside the spheres. Figure 3c-3f demonstrate that the generated WO3 nanoparticles are well dispersed in silica spheres for all doped SEP complexes after the calcination. Etching the sintered silica spheres with HF solution, isolated and narrow size distributed WO3 nanoparticles were found visibly, as shown in Figure 3g. The statistics for the nanoparticles in TEM view suggests the size of 2.2 nm on average (Figure S5b). Identically, we observed WO3 nanoparticles from other sintered samples, W-SEP-2/SiO2, W-SEP-3/SiO2, and W-SEP-4/SiO2 (Figure S4). Significantly, besides the encapsulation, the silica matrix around dispersed SEP complexes providing confined volume for decomposition of SEPs is very important to get the well dispersed WO3 nanoparticles, because we can not obtain well dispersed nanosized objects just through sintering the SEP complexes alone, as seen in the case of W-SEP-1 (Figure 3h). Porous Structure Determination of WO3-Doped Silica Spheres. The SEP-doped silica spheres before and after calcination were evaluated by gas adsorption. For pure silica spheres, S-0 shows no obvious pore structure, whereas W-0 reveals a 0.8 nm micropore existence, corresponding to the inherent microporous structure of silica, due to heating induced wiping off of the stuffing blocked in microporous structure. In contrast to W-0, for SEP-doped silica spheres, in addition to a similar microporous structure, a mesoporous structure was also found in W-SEP-1/ SiO2. The observed mesopores have a size region of 2.0-2.5 nm and show no distinct change with increase in the content of SEP from W-1 to W-5 (Table S4). From comparison of W-1, W-3, W-SEP-2/SiO2, and W-SEP-3/SiO2 (Figure 7), the pore diameter is close to the size of the SEP complexes under the employed CTAB ratio and increases from 2.5 nm for W-2 to 2.9 nm for W0 -2 when increasing the content of CTAB (Figure S7 and Tabel S4). Langmuir 2010, 26(18), 14894–14900

Article

Therefore, we believe that it is the SEP complexes contributing to the templates for the formation of mesopores. The microporous and mesoporous structures emerging in silica spheres are significant for the functionalization of the hybrid materials, because substance transportation in porous structures is favorable to the applications such as catalysis and gas sensor of the generated WO3 nanoparticles. The detailed XRD characterization reveals a halo in the region of 1-3° (2q), indicating an anomalous porous structure in silica spheres (Figure S8). Specific surface area measurement provides consistent results with the porous structure examination (Table S4). Different from the virgin SEP-1/SiO2 spheres that have small BrunauerEmmet-Teller (BET) surface areas between 4.6-16.8 m2/g, which show a decreasing tendency from S-1 to S-5, the corresponding sintered samples exhibit BET surface areas of 184.4395.5 m2/g, revealing a large increasing in the specific surface area. Meanwhile, from W-SEP-1/SiO2 (W-3) to W-SEP-2/SiO2 and W-SEP-3/SiO2 with the similar SEP doping content in silica spheres, the specific surface areas of sintered samples increase from 326 to 410 and 527 m2/g as well. Increase of CTAB ratio from W-2 to W-20 also induces the increase in surface area from 205 to 459 m2/g. All of these regular results can be well explained when the doped SEP complexes are regarded as the template of porous structure; that is, the high doping content, more alkyl chains covered on POM, and CTABs that cover on SEP complexes lead to the same tendency, the increase in specific surface area. Photovoltaic Property of WO3 Nanoparticle Incorporated Silica Spheres. As the UV-visible DRS spectrum performs a useful spectroscopic technique for solid samples, we employed the method to detect the bandgap of obtained WO3 incorporated in silica spheres. The position of the absorption edge is sensitive to the bonding between metal oxide polyhedron units,24 and thus, it has been used to quantitatively characterize the average particle size of nanocrystalline insulators and semiconductors.25 The spectra in Figure 8a show that the obtained nanostructured WO3 derived from pure SEP-1 and S-5 exhibit stronger absorption at wavelengths shorter than the bandedge cutoff of commercial WO3 reported in the literature.7d The cutoffs estimated from the Tauc plots afford the optical bandgap of 2.60 and 2.69 eV, larger but approaching the value (2.59 eV) of monoclinic WO3.26 In comparison to the nanostructured WO3 of W-SEP-1, the isolated small sized WO3 nanoparticles reveal a bit larger bandgap value, which also happened to WO3 from SEP2 and SEP-3 sintered in silica and nature, respectively. For WO3 nanoparticles prepared from other sintered SEP complexes, as shown in Figure 8, W-SEP-2/SiO2 and W-SEP-3/SiO2 spheres show bandgap values of 2.63 and 2.72 eV, respectively, whereas W-SEP-4/SiO2 spheres that contain pure WO3 nanoparticles gives the value of 2.42 eV. We believe that the bandgap difference among W-SEP/SiO2 should source from the hetero element incorporated in WO3 nanocrystalline, although no evidence can be provided through experiment in detail at present stage, due to the much less heterocomposition existing in POM-1, POM-2, and POM-3. Apparently, the higher bandgaps of WO3 nanoparticles (W-SEP/SiO2) than those of the nanostructured WO3 (W-SEP), except for the case of SEP-4, which presents similar value, should be due to size effect and the environment of silica matrix. To study the photoresponse of the obtained nanomaterials, their photovoltaic properties were characterized by SPV measurements.18 The WO3 nanoparticles (W-5) prepared from (24) Weber, R. S. J. Catal. 1995, 151, 470–474. (25) Brus, L. E. J. Chem. Phys. 1984, 80, 4403–4409. (26) Barton, D. G.; Shtein, M.; Wilson, R. D.; Soled, S. L.; Iglesia, E. J. Phys. Chem. B 1999, 103, 630–640.

DOI: 10.1021/la102491a

14899

Article

Zhao et al.

Figure 8. UV-DRS spectra of (a) W-5 and W-SEP-1, (b) W-SEP-2/SiO2, W-SEP-3/SiO2, and W-SEP-4/SiO2, and (c) W-SEP-2, W-SEP-3, and W-SEP-4, insets correspond to Tauc plots obtained from UV-DRS transform.

Figure 9. SPV spectra of pure WO3 and POM-1, W-SEP-1, W-5, and W-SEP-4/SiO2.

the SEP complexes reveal clear photovoltaic responses which become weaker with decreasing loading content of SEPs preincorporated in silica spheres (Figure 9), and n-type semiconductor property which is consistent with published results. The photovoltaic response is weak for the virgin POM-1 powder, whereas it increases evidently for sintered SEP-1, and the photovoltaic response region extends from ca. 390 nm for POM-1 to 465 nm around, which falls in the similar region of the UV-visible absorption of these species. Interestingly, the sintered SEP doped silica spheres, such as W-5, display strongly increased photovoltaic response intensity. Evidently, the photovoltaic response of these hybrid nanomaterials is derived from the WO3 nanoparticles, because a similar response is also found in both cases of sintered SEP-4/SiO2, which has no other metal oxide compositions, and monoclinic crystalline WO3 prepared by H2WO4 from sol-gel reaction.

Conclusion On the basis of the present strategy, we have successfully transferred SEP complexes, which are employed as the tungsten source due to the composition, into aqueous solution through a convenient phase-transfer approach by using CTAB that inserts its hydrophobic tails into the organic surface of SEPs. Such a route not only protects the frame structure of POMs from the chemical microenvironment during the sol-gel process but also increases the loading content of SEP complexes in the silica matrix efficiently, which helps to promote effective functionalization of silica particles. The doping content of SEP complexes does

14900 DOI: 10.1021/la102491a

not affect the well dispersing state of SEPs in silica spheres. More importantly, well separated small sized monoclinic crystalline WO3 nanoparticles with a narrow distribution are in situ generated due to the good dispersion of SEP complexes in silica spheres. As anticipated, the dispersed naked WO3 nanoparticles located in cavities in the silica matrix reveal higher band gaps due to size effect and visible photovoltaic properties. In addition, the SEP complexes are found serving as the templates of the mesoporous structure. The increase in specific surface area of hybrid silica spheres can improve the performance of WO3 nanoparticles in potential applications such as catalysis and gas sensor, as it provides an efficient transport pathway for mass exchange in the reaction. Finally, the present research demonstrates a general route for preparation of doped WO3 nanoparticles by controlling the chemical composition of POMs with different metal substituted polyoxotungstates. The whole procedure should be generally suitable for fabricating other SEP/SiO2 particles and corresponding silica particle pore supported metal oxide nanoparticles. Acknowledgment. This work was financially supported by National Basic Research Program of China (2007CB808003), National Natural Science Foundation of China (20703019, 20731160002, 20973082, and 0973042), 111 Project (B06009) for supporting the fruitful discussion with Prof. I. Kim at Pusan National University, and open project of State Key Laboratory of Polymer Physics and Chemistry of CAS. Supporting Information Available: Figures showing The TGA data of SEP-2, SEP-3, and SEP-4; IR spectra and assignments of POM-2, POM-3, POM-4, SEP-2, SEP-3, SEP-4, SEP-2/SiO2, SEP-3/SiO2, and SEP-4/SiO2; ζ potential of CTAB/SEP-1 and S1; statistical size distribution of S1 and WO3 from W-5; SEM of S-2, S-3, S-4, and S-5; TEM of W-SEP-2/SiO2, W-SEP-3/SiO2, WO3 from W-SEP-3/SiO2, and WO3 from W-SEP-4/SiO2; XRD data and Raman spectra of W-SEP-2/SiO2, W-SEP-3/SiO2, and W-SEP-4/SiO2; BET and pore dispersion of all samples; SXRD of W-5; and EA of S-2, SEP-2/SiO2, SEP-3/SiO2 and corresponding W-2, W-SEP-2/SiO2, and W-SEP-3/SiO2. This material is available free of charge via the Internet at http:// pubs.acs.org.

Langmuir 2010, 26(18), 14894–14900