Chemical Growth and Photochromism of Silver Nanoparticles into a Mesoporous Titania Template

pubs.acs.org/Langmuir © 2009 American Chemical Society

Chemical Growth and Photochromism of Silver Nanoparticles into a Mesoporous Titania Template Laurence Bois,*,† Fernand Chassagneux,† Yann Battie,‡ Franc- ois Bessueille,§ Laurent Mollet,† St
ephane Parola,† Nathalie Destouches,‡ Nelly Toulhoat,^,# and Nathalie Moncoffre^ † Laboratoire des Multimat
eriaux et Interfaces, UMR CNRS 5615, B^ at. Berthollet, Universit
e Claude Bernard, Lyon 1, 43 Boulevard 11 novembre 1918, 69622 Villeurbanne, France, ‡Laboratoire Hubert Curien, CNRS UMR 5516, Universit
e Jean Monnet, 18 Rue Prof. B. Lauras, Bat F, F-42000 Saint-Etienne, France, §Laboratoire des Sciences Analytiques, UMR 5180 CNRS, B^ at. Raulin, Universit
e Lyon 1, 43 Bd 11 novembre 1918, 69622 Villeurbanne, France, ^Universit
e Lyon 1, IPNL, CNRS/IN2P3, 4 rue E. Fermi, 69622 Villeurbanne cedex, France and #CEA/DEN, Centre de Saclay, 91191 Gif sur Yvette Cedex, France

Received June 30, 2009. Revised Manuscript Received August 27, 2009 Elaboration of mesoporous titanium oxide film supporting silver nanoparticles is described. Mesoporous titanium oxide films are characterized by TEM analysis. Titania films are infiltrated with a silver salt solution and chemical reduction treatments are performed using either a NaBH4 or a formaldehyde solution. Infiltrated films are then characterized by TEM, SEM, AFM, UV-visible spectroscopy, X-ray diffraction, and Rutherford Backscattering Spectrometry (RBS). The utilization of a mesoporous titania substrate allows to control the nanoparticle size and the interparticle distance. RBS experiments provide the evidence that NaBH4 treatment induces a strong accumulation of silver nanoparticles in the subsurface of the layer, while formaldehyde treatment induces the formation of silver nanoparticles embedded into almost the whole depth of the titania film. Large silver nanocrystals are also formed at the film surface whatever the reducer used. A broad visible absorption band related to the surface plasmon resonance (SPR) is obtained in both cases and is strongly red-shifted compared to the SPR obtained for silver nanoparticles inside a silica matrix. Moreover, irradiation with visible light causes the photooxidation of silver nanoparticles by titania and a complete discoloration of the material. The photooxidation is related to a drastic decrease in the silver nanoparticle size and is found to be reversible, particularly in the case of the material obtained by the formaldehyde reduction.

1. Introduction Noble metal nanoparticles and especially silver nanoparticles exhibit an absorption band in the visible region caused by the surface plasmon resonance (SPR); as well-known the SPR wavelength and intensity depend on the nanoparticle size, geometry and environment (dielectric constant and interparticle distance).1-3 The photochromism of silver species adsorbed on colloidal titania has been known for a long time.4-6 A darkening of these systems upon illumination was found to be reversible and the oxidation of silver species was proven to explain the bleaching in these systems. By UV irradiation, electrons at the valence band of titania are excited to the conduction band and then migrate to Ag(I) species reducing them to Ag(0). Silver nanoparticles interact with visible light through their surface plasmon resonance, and their electrons migrate to the conduction band of titania, which induces the Ag(0) oxidation into Ag(I) species. Recently, this property has been revisited and a new functionality has been reported, for nanocomposites constituted of silver nanoparticles *To whom correspondence should be addressed. E-mail: laurence.bois@ univ-lyon1.fr. Phone: 33 04 72 43 14 00. Fax: 33 04 72 44 06 18. (1) El-Sayed, M. A. Acc. Chem. Res. 2001, 34(4), 257–264. (2) Faraday, M. Philos. Trans. R. Soc. 1857, 147, 145. (3) Mulvaney, P. Langmuir 1996, 12, 788–800. (4) Goetz, A.; In, E. V. C. Rev. Mod. Phys. 1948, 20, 131–142. (5) Clarks, W. C.; Vondjdis, A. Nature 1964, 203, 635–636. (6) Fleischauer, P. D.; Alan Kan, H. K.; Shepard, J. R. J. Am. Chem. Soc. 1972, 94, 283–285. (7) Ohko, Y.; Tatsuma, T.; Fujii, T.; Naoi, K.; Niwa, C.; Kubota, Y. Nat. Mater. 2003, 2(1), 29–31.

Langmuir 2010, 26(2), 1199–1206

embedded in a nanoporous titanium oxide film, namely the multicolor photochromism.7-16 The color of these composites can be indeed modified through irradiation with a monochromatic visible light and can be reversibly erased by UV illumination (the color of the irradiation light is then replaced with a brown color). The multicolor photochromism property opens the application field to optical data storage, display devices, and so on 16. On the other hand, titania is used in a lot of applications, e.g., in solar cells,17 sensors,18 batteries,19 and in photocatalysis20-22 and a lot of studies are also devoted to the addition of noble metals to (8) Naoi, K.; Ohko, Y.; Tatsuma, T. J. Am. Chem. Soc. 2004, 126(11), 3664– 3668. (9) Naoi, K.; Ohko, Y.; Tatsuma, T. Chem Commun. 2005, 10, 1288–1290. (10) Matsubara, K.; Kelly, K. L.; Sakai, N.; Tatsuma, T. Phys. Chem. Chem. Phys. 2008, 10(16), 2263–2269. (11) Matsubara, K.; Tatsuma, T. Adv. Mater. 2007, 19(19), 2802–2806. (12) Kelly, K. L.; Yamashita, K. J. Phys. Chem. B 2006, 110, 7743–7749. (13) Kawahara, K.; Suzuki, K.; Ohkowa, Y.; Tatsuma, T. Phys. Chem. Chem. Phys. 2005, 7, 3851–3855. (14) Dahmen, C.; Sprafke, A. N.; Dieker, H.; Wuttig, M.; von Plessen, G. Appl. Phys. Lett. 2006, 88, 011923-1–3. (15) Okumu, J.; Dahmen, C.; Sprafke, A. N.; Luysberg, M.; Von Plassen, G.; Wuttig, M. J. Appl. Phys. 2005, 97, 094305. (16) Qiao, Q.; Zhang, X.; Lu, Z.; Wang, L.; Liu, Y.; Zhu, X.; Li, J. Appl. Phys. Lett. 2009, 94, 074104-1–3. (17) Gr€atzel, M. Coord. Chem. Rev. 1991, 111, 167–174. (18) Iwanaga, T.; Hyodo, T.; Shimizu, Y.; Egashira, M. Sens. Actuators B 93 2003, 1-3, 519–525. (19) Long, J. W.; Dunn, B.; Rolison, D. R.; White, H. S. Chem. Rev. 2004, 104 (10), 4463–92. (20) Fox, M. A. Acc. Chem. Res. 1983, 16, 314–21. (21) Jakob, M.; Levanon, H.; Kamat, P. V. Nano Lett. 2003, 3, 353–358. (22) Patel, M. N.; Williams, R. D.; May, R. A.; Uchida, H.; Stevenson, K. J.; Johnston, K. P. Chem. Mater. 2008, 20(19), 6029–6040.

Published on Web 09/18/2009

DOI: 10.1021/la902339j

1199

Article

TiO2 as photocatalyst21-33 or as antimicrobial film.28,34,35 Generally, the silver nanoparticles growth in a titania matrix is performed by UV irradiation,36-38 by chemical reduction, for instance, with NaBH4,39,40 or recently with the classical silver mirror reaction.41 The design of new composites with silver nanoparticles embedded in a titania matrix with a controlled porosity could lead to a tailoring of the properties, in particular of the optical characteristics, because the matrix porosity may act as a template in the silver nanoparticle growth. Mesoporous materials are characterized by a high specific surface area and controlled pore diameter.42-46 Mesoporous titania films are prepared by using structure-directing block copolymers and the evaporation-induced self-assembly strategy.45,46 Such films, which have very high specific surface area and a controlled porosity in the range 5-15 nm, are often considered as a way to adjust electronic, optical, or catalytic properties.43-47 Mesoporous titania has already been used to grow metallic nanoparticles within their porosity, classically by reduction of (23) Kamat, P. V. J. Phys. Chem. C 2007, 111(7), 2834–2860. (24) Wood, A.; Giersig, M.; Mulvaney, P. J. Phys. Chem. B 2001, 105(37), 8810– 8815. (25) Chan, S. C.; Barteau, M. A. Langmuir 2005, 21, 5588–5595. (26) Ohtani, B.; Zhang, S.-W.; Ogita, T.; Nishimoto, S.; Kagiya, T. J. Photochem. Photobiol., A: Chem. 1993, 71, 195–198. (27) Awazu, K.; Fujimaki, M.; Rockstuhl, C.; Tominaga, J.; Murakami, H.; Ohki, Y.; Yoshida, N.; Watanabe, T. J. Am. Chem. Soc. 2008, 130, 1676–1680. (28) Page, K.; Palgrave, R. G.; Parkin, I. P.; Wilson, M.; Savin, S. L. P.; Chadwick, A. V. J. Mater. Chem. 2007, 17, 95–104. (29) Zhang, L.; Yu, J. C.; Yin Yip, H.; Li, Q.; Kwong, K. W.; Xu, A.-W.; Wong, P. K. Langmuir 2003, 19, 10372–10380. (30) Andersson, M.; Birkedal, H.; Franklin, N. R.; Ostomel, T.; Boettcher, S.; Palmqvist, A E. C.; Stucky, G D. Chem. Mater. 2005, 17, 1409–1415. (31) Hirakawa, T.; Kamat P. V. Langmuir 2004, 6, 20, 14, 5645-5647. (32) Yates, H. M.; Brook, L. A.; Sheel, D. W. Int. J. Photo. 2008, ID 1870392. (33) Jin, M.; Zhang, X.; Nishimoto, S.; Liu, Z.; Tryk, D.; Emeline, A.; Murakami, A. V.; Fujishima, T. A. J. Phys. Chem. C 2007, 111, 658–665. (34) Sharma, V. K.; Yngard, R. A.; Lin, Y. Adv. Colloid Interface Sci. 2009, 145, 83–96. (35) Gunawan, C.; Teoh, W. Y.; Marquis, C. P.; Lifia, J.; Amal, R. Small 2009, 5, 341–344. (36) Cozzoli, P. D.; Comparelli, R.; Fanizza, E.; Curri, M. L.; Agostiano, A.; Laub, D. J. Am. Chem. Soc. 2004, 126, 3868–3879. (37) Paramasivam, J. M.; Macak; Ghicov, A.; Schmuki, P Chem. Phys. Lett. 2007, 445(4-6), 233–237. (38) Zhang, H.; Wang, G.; Chen, D.; Lv, X.; Li, J. Chem. Mater. 2008, 20, 6543– 6549. (39) Nino-Martinez, N.; Martinez-Castanon, G. A.; Aragon-Pina, A.; Martinez-Gutierrez, F.; Martinez-Mendoza, J. R.; Ruiz, F. Nanotechnology 2008, 19, 065711/1–065711/8. (40) He, J.; Ichinose, I.; Kunitake, T.; Nakao, A. Langmuir 2002, 18, 10005– 10010. (41) Shan, Z.; Wu, J.; Xu, F.; Huang, F.-Q.; Ding, H. J. Phys. Chem. C 2008, 112, 15423–15428. (42) Kresge, C. T.; Leonowicz, M. E.; Roth, W. K.; Vartulim, J. C.; Beck, J. S. Nature 1992, 395, 710–712. (43) Grosso, D.; de Soler-Illia, Galo J. A. A.; Babonneau, F.; Sanchez, C.; Albouy, P.-A.; Brunet-Bruneau, A.; Balkenende, A. Ruud Adv. Mater. 2001, 13, 1085–1090. (44) Crepaldi, E. L.; de Soler-Illia, Galo J.; Grosso, D.; Cagnol, F.; Ribot, F.; Sanchez, C. J. Am. Chem. Soc. 2003, 125, 9770–9786. (45) Grosso, D.; de Soler-Illia, G. J. A. A.; Crepaldi, E. L.; Cagnol, F.; Sinturel, C.; Bourgeois, A.; Brunet-Bruneau, A.; Amenitsch, H.; Albouy, P A.; Sanchez, C. Chem. Mater. 2003, 15, 4562–4570. (46) Koh, C.-W.; Lee, U-H; Song, J-K; Lee, H-R; Kim, M-H; Suh, M.; Kwon, Y-U Chem. Asian J. 2008, 3, 862–867. (47) Liu, K.; Zhang, M.; Shi, K.; Fu, H. Mater. Lett. 2005, 59(26), 3308–3310. (48) Stathatos, E.; Choi, H.; Dionysiou, D. D. Environ. Eng. Sci. 2007, 24, 13– 20. (49) Wang, D.; Zhou, F.; Wang, C.; Liu, W. Microporous Mesoporous Mater. 2008, 116, 658–64. (50) Stathatos, E.; Lianos, P.; Falaras, P.; Siokou, A. Langmuir 2000, 16, 2398– 2400. (51) Wang, X.; Yu, J. C.; Ho, C.; Mak, A. C. Chem. Commun. 2005, 2262–2264. (52) Liu, Y.; Wang, X.; Yang, F.; Yang, X. Microporous Mesoporous Mater. 2008, 114(1-3), 431–439. (53) Yu, J. C.; Wang, X.; Wu, L.; Ho, W.; Zhang, L.; Zhou, G. Adv. Funct. Mater. 2004, 14, 1178–1183. (54) Perez, M. D.; Otal, E.; Bilmes, S. A.; Soler-Illia, G. J. A. A.; Crepaldi, E. L.; Grosso, D.; Sanchez, C. Langmuir 2004, 20, 6879–6886.

1200 DOI: 10.1021/la902339j

Bois et al.

an ionic precursor inside the pores by UV radiation,48-53 through an electrochemical process,22,54 a thermal route,30,55 or very recently, by using the formaldehyde process.56 Besides the enhanced catalytic properties and the antimicrobial activity,52 new physical properties are expected from the interaction between Ag and TiO2,51 for instance, for single molecule spectroscopy and biological objects visualization.57-59 In this paper, we were interested in the elaboration of new nanocomposites constituted of silver nanoparticles embedded in a mesoporous titania film. Very few studies have already been related to the chemical growth of silver nanoparticles into these kind of templates.56 A mesoporous TiO2 matrix has been prepared from a triblock copolymer P123 (PEO20-PPO70-PEO20). Silver nanoparticles have then been grown into the mesoporous titania using a two-step process: namely, infiltration of a silver salt and then chemical reduction involving sodium borohydride or formaldehyde reducing agents. Finally, the surface plasmon resonance (SPR) and the possibility to vanish the SPR by visible illumination has been qualitatively studied.

2. Experimental Section 2.1. Materials. Silver nitrate (AgNO3), sodium borohydride (NaBH4), formaldehyde (HCHO), tetraorthobutyltitanate (Ti(OBu)4), and block copolymer P123 ((PEO)20(PPO)70(PEO)20) were all purchased from Aldrich. 2.2. Mesoporous Titania Films. For the synthesis of mesoporous titania films, a method involving the evaporation-induced self-assembly (EISA) route, the nonionic amphiphilic triblock copolymer P123 as the structurating agent and tetrabutylorthotitanate (TBT) as titanium oxide precursor was used.47 A solution of P123 (1 g) in ethanol (12 g) was added to a solution of TBT (3.4 g) in concentrated HCl (3.2 g). Films were prepared by dipcoating on cleaned glass slides. A 20 mm/min dip-coating rate was used. After a 24 h drying step at room temperature, films were calcined at 350 C during 4 h with a heating rate of 1 C/min. 2.3. Silver Nanoparticles in Mesoporous Films. Mesoporous titania films were immersed during 30 min in an aqueous ammoniacal silver solution [Ag(NH3)2]NO3, obtained as follows. At first a silver nitrate solution was prepared at a concentration fixed at 50 mM; a NH3 solution was then added until a clear solution is obtained. Immediately after, the substrate was immersed in a freshly prepared NaBH4 aqueous solution (50 mM) for 30 s, rinsed in water, and air-dried at 100 C. The reduction of silver by NaBH4 occurs very rapidly, as is shown by the immediate film coloration. To limit the loss of silver nanoparticles toward the solution, shown by its progressive darkening, it is important to use very short reaction times in the NaBH4 solution. In another case, the titania substrate containing ammoniacal silver salt was immersed in an ethanolic solution (27 mL) containing formaldehyde (0.55 mL) and water (1.95 mL) and heated at 80 C for 30 min. 2.4. Characterization. The TEM images were taken with a TOPCON EM002B transmission electron microscope operating at 200 kV. Samples for the transmission electron microscopy (TEM) characterization were prepared by scrapping the film. The fragments were then deposed directly on a copper grid coated with holey carbon film. UV-visible absorption spectra were recorded on a Perkin elmer lambda 35. X-ray diffraction (XRD) patterns were recorded on a Philips Xpert Pro diffractometer equipped (55) Li, X. S.; Fryxell, G. E.; Wang, C.; Engelhard, M. H. Microporous Mesoporous Mater. 2008, 111, 639–642. (56) Fuertes, M. C.; Marchena, M.; Marchi, M. C.; Wolosiuk, A .; Soler-Illia, G. J. A. A. Small 2009, 2, 272–282. (57) Zolotavin, P.; Permenova, E.; Sarkisov, O.; Nadtochenko, V.; Azouani, R.; Portes, P.; Chhor, K.; Kanaev, A. Chem. Phys. Lett. 2008, 457, 342–346. (58) Aiboushev, A. V.; Astafiev, A. A.; Lozovik, Y. E.; Merkulova, S. P.; Nadtochenko, V. A.; Sarkisov, O. M. Phys. Lett. A 2008, 372, 5193–5197. (59) Martinez, E. D.; Bellino, M. G.; Soler-Illia, G. J. A. A. Appl. Mater. Interfaces 2009, 1, 746–749.

Langmuir 2010, 26(2), 1199–1206

Bois et al.

Article

Figure 1. TEM images of titania film without silver nanoparticles: (a) top-view, (b) cross-section. TEM images of titania film with silver nanoparticles from NaBH4 reduction: (c) top-view, (d) cross-section, (e) high resolution. with a monochromator, using Cu KR radiation. Rutherford backscattering spectrometry was performed on the 4 MV Van de Graaff accelerator of the IPNL (Institut de Physique Nucl
eaire de Lyon), delivering a 2 MeV helium beam. The Heþ backscattered particles were detected at a 172 angle. The beam intensity was set at 22 nA. The SIMNRA program was used to simulate the energy spectra and to determine the atomic concentration profiles.60 XPS analysis (X-ray photoelectron spectroscopy) was performed on a PHI Quantera SXM. Atomic Force microscopy (AFM) scans were recorded on a 5500 Picoplus (Agilent Technologies) instrument equipped with a Si tip and operating in tapping mode. SEM images were acquired on a Hitachi S800 equipment, at 15 kV after the films are covered with a gold-palladium layer. Irradiations are performed with a “white light” using an halogen lamp (P = 250 W) during 4 h or with a green laser (Melles Griot) (P = 5mW; λ= 543.5 nm) or a red laser (Melles Griot) (P = 15mW; λ= 632.8 nm) during 15 min.

3. Results 3.1. Characterization of the Titania Film. Amorphous titania films are obtained as described in the Experimental Section. The TEM image of an empty mesoporous titania film, before the silver nanoparticle growth, is shown in Figure 1. The top-view image (Figure 1a) is coherent with a disordered, wormlike mesostructure.61-63 Pores display a characterized size and interdistance. This wormlike phase is formed by a disordered stacking of spherical pores (probably deriving from the bodycentered cubic mesophase Im3m).44,45 The pore to pore distance can be estimated from this view to about 11 nm. The pore size is around 7-8 nm. The wall thickness is approximated between 2 and 3 nm. On the cross-section view (Figure 1b), the contraction along the direction perpendicular to the substrate, already ob(60) Mayer, M. SIMNRA User’s guide, Technical Report IPP 9/113; Max Planck Institut f€ur Plasmaphysik: Garching, Germany, 1997. (61) Pan, J. H.; Lee, W. I. New. J. Chem . 2005, 29, 841–846. (62) Nicole, L.; Boissiere, C.; Grosso, D.; Quach, A.; Sanchez, C. J. Mater. Chem. 2005, 15, 3598–3627. (63) Sanchez, C.; Boissiere, C.; Grosso, D.; Laberty, C.; Nicole, L. Chem. Mater. 2008, 20, 682–737.

Langmuir 2010, 26(2), 1199–1206

served in these kind of materials,45 is clearly seen. The pore dimension along the direction perpendicular to the substrate is around 2 and 3 nm. 3.2. Characterization of the Titania Film with Silver Nanoparticles Grown with NaBH4. Silver nanoparticles are grown inside the pores of the mesoporous titania film by reduction in a sodium borohydride solution after immersion into a [Ag(NH3)2]þ solution. The TEM images of silver nanoparticles grown in the mesoporous titania film are presented in Figure 1. On the top view (Figure 1c), the surface is covered of silver nanoparticles with diameter in the 15-20 nm range. Moreover, the porous film is filled with silver nanoparticles with diameter between 6 and 8 nm. On the TEM picture which probably correspond to the “cross-section” of the film (Figure 1d), it seems that silver nanoparticles are concentrated on the subsurface of the film. But the strong evidence of this fact will be provided later by the RBS experiment. On the HR-TEM picture (Figure 1e), polycrystalline silver particles of diameter 21 and 14 nm, exhibiting moires, are observed on the surface of the film. Smaller monocrystalline silver particles are observed in the deepness of the titania layer, with diameter between 6 and 11 nm, and sometimes (111) lattice franges of cfc silver are seen (zoom in Figure 1e). The distance between two silver particles is very short, a few nanometers, which is probably the wall thickness of the mesoporous titania. SEM and AFM images are shown in Figure 2. From theses images (Figure 2), different population of silver crystals is observed between 20 and 100 nm. These silver nanocrystals are probably formed by diffusion and aggregation of the smaller ones out the titania template. The film gets a light purple coloration. The visible absorption spectrum (Figure 3a) shows a broadband that consists of at least two bands, the first at about 450 nm and the second at 580 nm. This absorption is related to the surface plasmon resonance of the silver nanoparticles formed in the titania substrate. Roughly, the absorption around 450 nm could probably be related to small silver nanoparticles, observed by TEM DOI: 10.1021/la902339j

1201

Article

Bois et al.

Figure 2. (a) AFM and (b) SEM images of films obtained with NaBH4 process.

Figure 3. UV-vis absorption spectra of films (a) after a NaBH4 process, (b) after formaldehyde process, (c) after formaldehyde and white light treatment, (d) after formaldehyde, white light, and formaldehyde. Photographs of films (e) after formaldehyde process, (f) after a white light treatment, (g) after a second formaldehyde process. XRD patterns of films (h) after NaBH4, (i) after formaldehyde, (j) after formaldehyde and white light illumination. (*): localization of the (111) fcc silver diffraction peak. () localization of the (111) Ag2O diffraction peak.

inside the porous film, whereas the absorption around 580 nm could be explained by the bigger one, observed by SEM and AFM outside the porous film. 1202 DOI: 10.1021/la902339j

The XRD experiment (Figure 3h) confirms that metallic silver is formed since a diffraction peak at 38.1 (fwhm = 0.35) may be attributed to the (111) Bragg’s reflection of the face-centered Langmuir 2010, 26(2), 1199–1206

Bois et al.

Article

Figure 5. TEM images of films (a) after formaldehyde process, top-view; (b) high-resolution; (c) cross-section. Figure 4. RBS patterns and silver profiles (inset) calculated from RBS patterns of films (a) after impregnation with silver salts, (b) after NaBH4 process, (c) after formaldehyde process.

cubic structure of silver phase metal. The average crystallite size is estimated to 25 nm, from the broadening of the (111) diffraction peak applying the Scherrer formula D = 0.94λ/fwhm cos θ. In this equation, D is the average crystallite size, λ is the X-ray wavelength, and θ the Bragg angle. This observation is in agreement with the TEM observation, which shows some particles with a size around 20 nm at the film surface. Experimental RBS spectra are presented in Figure 4. The spectrum of the sample before the silver reduction is shown in Figure 4a, whereas the spectra of the samples after silver reduction respectively in NaBH4 or in formaldehyde are shown in Figure 4b and 4c respectively. The spectrum of the sample after silver salt impregnation and before any reduction consists of a broad silver peak (Figure 4a). After reduction with NaBH4, the silver signal is significantly modified (Figure 4b). The silver peak becomes narrower, which reveals that a silver depletion occurs in the depth of the film. The silver profiles as function of depth are presented in Figure 4 (insets). Silver profile before silver reduction is represented in Figure 4a (inset), whereas silver profiles after reduction with NaBH4 or formaldehyde, respectively, are shown in the insets to panels b and c in Figure 4. They are deduced from the RBS spectra, using the SIMNRA software. These simulations allow measuring a silver atomic concentration of 7.2% and a Langmuir 2010, 26(2), 1199–1206

titanium atomic concentration of 25.1%, leading to the ratio Ag/Ti = 0.29 for the sample before silver reduction. The calculated film thickness is 120 nm. As suggested by the TEM observation, the RBS analysis provides, for the first time as far as we know, a strong evidence of the metal repartition along a mesoporous layer. Indeed, after the NaBH4 reduction, most of silver (62%) is concentrated in the first 15 nm of the layer (Figure 4b). Similar observation is reported earlier when growing Ag nanoparticles with NaBH4 in a mesostructured composite silica-copolymer film.64,65 To confirm the high Ag/Ti ratio measured with RBS analysis, we have performed an XPS analysis of the mesoporous film after the formation of silver nanoparticles. The following atomic percents have been measured: C, 19.5%; O, 54.0%; Ti, 19.8%; Ag, 6.7%. The ratio Ag/Ti equal to 0.34, calculated from XPS measurement, is consistent with the RBS result. 3.3. Characterization of the Titania Film with Silver Nanoparticles Grown with Formaldehyde. The influence of the reductive treatment is observed by replacing the sodium borohydride reducer by another reactant reducer: the formaldehyde. In this case, a higher density of silver nanoparticles is observed on the TEM image (Figure 5a), where spherical particles (64) Bois, L.; Chassagneux, F.; Parola, S.; Bessueille, F.; Battie, Y.; Destouches, N.; Boukenter, A.; Moncoffre, N.; Toulhoat, N. J. Solid State Chem. 2009, 182, 1700–1707. (65) Bois, L.; Bessueille, F.; Chassagneux, F.; Battie, Y.; Destouches, N.; Hubert, C.; Boukenter, A.; Parola, S. Colloids Surf. 2008, 325, 86–92.

DOI: 10.1021/la902339j

1203

Article

Bois et al.

Figure 6. AFM and SEM (inset) images of films obtained (a) with formaldehyde process and (b) with formaldehyde process after white light treatment.

with diameter between 5 and 11 nm are seen. On the highresolution TEM image (Figure 5b), silver nanoparticles with diameter of 7 and 10 nm are noted. The space between two particles is typically in the 2-3 nm range. The cross-section presented in Figure 5c shows that silver nanoparticles are located within the entire depth of the film. SEM and AFM images are shown in Figure 6a. From the AFM image (Figure 6a), a population of silver crystals is observed with sizes around 30 nm. The same observation may be done from the SEM image (Figure 6a insert), because the surface is covered with particles with sizes between 20 and 30 nm. According to the high density of nanoparticles evidenced by TEM analysis, a very deep purple coloration is obtained (Figure 3e). The absorption spectrum shows a broadband at 510 nm with a high absorbance (1.4) (Figure 3b). A second absorption band around 450 nm is also probably formed. Once again, the absorption band at lower wavelength may be in a first approximation related to smaller silver nanoparticles, while those at higher wavelength could be related to bigger silver nanoparticles at the film surface. Using X-ray diffraction on the film, the silver metallic phase is observed as the (111) reflection at 37.9 (fwhm = 0.61) (Figure 3i). It may be related to smaller silver nanoparticles whose size can be estimated at 15 nm from the Scherrer relation. This observation is coherent with the TEM analysis, which shows particles with diameter between 5 and 10 nm. From RBS spectrum (Figure 4c), precious information is provided on the silver repartition through the titania layer, because it can be observed that after the reduction with formaldehyde, the silver distribution is not strongly changed, except a slight silver enrichment at the surface. The silver profile (Figure 4c) confirms that silver is more concentrated (38%) in the first 25 nm of the layer. Consequently, the silver repartition within the titania layer differ considerably according to the reducer used. 3.4. Influence of a Visible Light Treatment. When, a film is illuminated with a white lamp during one hour, the coloration vanished, as reported in Figure 3f and c. By using a laser red illumination (15 min), a red coloration of the material obtained with formaldehyde is observed (Figure 7a and b). By the same way, with a laser green illumination (5 min), a coloration between blue and green appears (Figure not shown). The multicolor photochromism in mesoporous titania films, recently described by Tatsuma et al.10 in the case of nanoporous titania films, will be the object of a further paper. 1204 DOI: 10.1021/la902339j

The color changes are attributed to surface plasmon resonance shifts which are related to the oxidation of silver nanoparticles.10 The XRD (111) reflection of fcc metallic silver (Figure 3j) is no more observable after the illumination time. A very weak peak is noted at 32.3 that can be tentatively assigned to the (111) reflection of cubic Ag2O (reported at 32.79) or to the (404) reflection of silver titanate Ag2Ti4O9 (reported at 32.17). AFM and SEM pictures in Figure 6b confirm that the presence of the large silver nanoparticles with size around 20-30 nm is strongly decreased after light illumination. On the AFM picture, some silver particles are still observed, but most of them have disappeared and the mesostructure of the titania film is clearly observed. To be sure that the light treatment does not induce a drastic change in the interaction between the AFM tip and the titania surface, the result obtained with AFM has been confirmed by SEM analysis. A TEM observation (Figure 7c) reveals the presence of silver nanoparticles with a diameter considerably decreased. When, the formaldehyde process is used, silver nanoparticles with diameter comprised between 3 and 4.5 nm are seen. Moreover, in some places of the matrix (see the square for instance in Figure 7d), some fringes are seen whose spacing has been measured at 0.289 nm. These fringes could be due to the presence of a silver titanate phase Ag2Ti4O9 (JCPDS 00-032-1028) because the (004) reflection is expected at 0.287 nm. For a film obtained from the NaBH4 reduction, similar observations have been done. A clear identification of some silver species is difficult because of their low crystallinity. 3.5. Influence of a Second Reductive Treatment. The reversibility of the process is shown through a second reductive treatment performed after the white illumination since the pictures of the films show that the purple coloration reappears (Figure 3g). With visible absorption spectroscopy, it is proved that the second formaldehyde process lead to an almost complete restoration of the purple color (Figure 3d). Nevertheless, some part of silver species may be lost during the reductive treatment. A SEM picture shows that after a second formaldehyde treatment, the large silver nanoparticles (with size about 50 nm) are seen on the surface film again (Figure 7e). Moreover, a TEM picture (Figure 7f) reveals the reconstruction of silver nanoparticles with diameter between 5 and 8 nm.

4. Discussion In our UV visible absorption results, a first component around 450 nm is observed as well as a second one at higher wavelength, Langmuir 2010, 26(2), 1199–1206

Bois et al.

Article

Figure 7. (a) Optical microscopy image and (b) photograph of the film after formaldehyde and a red laser treatment. TEM images of the film (c) after formaldehyde and white light illumination and (d) high resolution. (e) SEM and (f) TEM images of the film after formaldehyde, white light treatment, and a second formaldehyde treatment.

580 nm in the NaBH4 process, and 510 nm in the formaldehyde process. The first component at 450 nm should probably be related to the smaller silver nanoparticles, formed inside the mesoporosity, whose size is around 6 and 8 nm. The absorption band of small silver nanoparticles inside a mesoporous titania film has been reported experimentally around 430 nm.50,51 This attribution is coherent if we consider that silver nanocrystals are not embedded in a titania matrix but inside a titania pore containing air and water, which will reduce the effective indice of the medium. Indeed, The SPR band around 450 nm is not commonly reported in the literature and does not correspond to small silver nanoparticles embedded in a TiO2 matrix according to the Mie theory, which leads to a SPR band centered at 490 nm in that case. To simulate such a low SPR wavelength, one has to take into account the location of the silver nanoparticles in the pores of the TiO2 matrix. Using the generalized Mie theory for core-shell particles,66 we calculated the extinction cross-section of a silver nanoparticle growing in a pore of 8 nm diameter. In this modeling, the metal core and the pore are supposed to form a concentric spherical heterosystem. The silver core, of radius R, has a dielectric function εAg equal to the sum of the experimentally determined bulk silver dielectric function εexp67 and a term that takes into account the classical free path effect εAg ¼ εexp þ

ω2p ω2p
 -  ωðωþiΓ0 Þ ω ωþi Γ0 þ vRf

In this expression, ω, ωp = 8.98 eV, vf = 1.39  106 m s-1, and Γ0 = 0.02 eV are the electromagnetic wave frequency, the plasma (66) Kreibig, U.; Vollmer, M. Optical Properties of Metal Clusters; SpringerVerlag: Berlin, 1995; p 58 (67) Palik, E. D. Handbook of Optical Constants of Solids; Academic Press: New York, 1985.

Langmuir 2010, 26(2), 1199–1206

Figure 8. Theoretical extinction cross-section spectra of a silver nanoparticle of increasing radius R in an air spherical volume of 8 nm diameter inside a TiO2 matrix.

frequency, the Fermi velocity, and the bulk free-electron relaxation frequency, respectively. The dielectric functions of the pore and the TiO2 matrix are assumed to be independent of wavelength and are set, respectively, at 1 and 4. When varying R from 3 to 4 nm, the simulated SPR wavelength, corresponding to the wavelength for which the extinction cross-section spectrum reaches a maximum (Figure 8), increases from 385 to 490 nm. The presence of various nanoparticle sizes below the pore diameter can then lead to an inhomogeneous broadening of the SPR spectrum and to the presence of a maximum around 450 nm. The second component at 510 or 580 nm is related to the larger silver nanoparticles, whose size is between 15 and 25 nm. This affirmation relies on the observations of Matsubara and others authors, who have noted the growth of silver nanocrytals on titania substrate from 10 to 25 nm diameter in asssociation with a DOI: 10.1021/la902339j

1205

Article

shift of the surface plasmon resonance from 480 to 510 nm.10,52,56 The dipolar mode of the largest silver nanocrystals can also contribute for the higher wavelength of the spectrum.68 For instance, silver nanocubes of size of 80 nm in solution present 6 strong resonances, the lowest at 480 nm and the highest at 550 nm due to quadripolar and dipolar resonances, respectively.68 These large silver nanoparticles, observed both by AFM, SEM, and to a smaller extent TEM and XRD, are probably formed by diffusion out of the titania pores of the smaller nanoparticles and by their aggregation. So large red-shift compared to the results found in silica systems,64,65 with the typical plasmon peak of Ag nanoparticles at 400 nm, may be explained by the high refractive index of the TiO2 matrix and probably by the large size of silver nanoparticles. Moreover, the close contact of adjacent silver particles could also be invoked.56,69 So, it is certainly quite difficult to clearly correlate a frequency of the SPR to the silver nanocrystals size, because a lot of contributions whith contradictory effects are involved. By comparing the NaBH4 and the formaldehyde reduction, it is found that a high accumulation of silver nanoparticles occurs in the subsurface of the layer when the NaBH4 process is used, whereas silver nanoparticles are formed almost within the complete thickness of the titania layer when the formaldehyde process is used. This original result concerning the repartition of metallic nanoparticles within a mesoporous layer, has been clearly evidenced for the first time by means of the RBS experiment. The difference may be related to the redox potentials because E0(BH4-/H2) = -0.96 V at pH 10, whereas the E(HCOOH/ HCHO) = -0.36 V at pH 6,70 and to distinct kinetic behaviors. The reduction of silver ions by BH4- occurs very rapidly before any diffusion of BH4- inside the thickness of the layer; so silver nuclei are formed in the first second in the subsurface of the film and silver ions or silver clusters71 probably diffuse toward the nuclei already formed. These clusters are formed by silver atoms via condensation reactions; it should be noted that the nature of some of them have already been determined; they generally carry a positive charge, such as Ag2þ, Ag42þ, Ag9þ.72,73 In the mechanism proposed by Henglein, there is nucleation of silver atoms up to a critical size and then the growth occurs via the surface reduction of silver ions.73 Moreover, as the surface charge of the titania surface is strongly negative in a basic solution,74 an electrostatic repulsion of the titania surface on anionic borohydride species could also explain the difficulty for borohydride to (68) Tao, A.; Sinsermsuksakul, P.; Yang, P. Angew. Chem., Int. Ed. 2006, 45, 4597–4601. (69) He, J.; Ichinose, I.; Fujikawa, S.; Kunitake, T.; Nakao, A. Chem. Commun. 2002, 1910–1911. (70) Pourbaix, M. Atlas d’
equilibres e
lectrochimiques a 25C,; Gauthier-Villars: Paris, 1963. (71) Linnert, T.; Mulvaney, P.; Henglein, A.; Weller, H. J. Am. Chem. Soc. 1990, 112, 4657–4664. (72) Henglein, A.; Giersig, M. J. Phys. Chem. B 1999, 103, 9533–9539. (73) Finney, E. E.; Finke, R. G. J. Colloid Interface Sci. 2008, 317, 351–374. (74) Kosmulski, M. Adv. Colloid Interface Sci. 2002, 99, 255–264.

1206 DOI: 10.1021/la902339j

Bois et al.

diffuse into the layer. Meanwhile, the formaldehyde can not be repulsed from the titania surface and can diffuse into the layer. In the case of a formaldehyde reduction, the reaction is slower, and the diffusion of formaldehyde inside the thickness is possible. Consequently, silver nuclei are formed within the complete layer. The photooxidation of silver nanoparticles in the presence of titanium oxide has already been reported.4-16 The silver nanoparticles, because of the surface plasmon resonance effect, are excited by visible light and the photoexcited electrons may be transferred to titania.38 There is a partial oxidation of silver nanoparticles,4-16 manifested with the size reduction of nanoparticles which then become non resonant. The decrease of the mean size of the silver nanocrystals, has already been evidenced by X-ray diffraction14 or AFM.11 In our study, the decrease in the size of silver nanoparticles is confirmed by TEM analysis for the smaller nanoparticles and by the disappearance of the metallic silver peak with X-ray diffraction. Moreover, the large silver nanocrystals have disappeared, as this phenomenon is clearly observed both by SEM and AFM. When performing a second reductive treatment, we can observe by TEM the reappearance of small silver nanoparticles (with sizes around 6 and 8 nm) and by SEM the reappearance of large silver nanoparticles (with a size around 50 nm).

5. Conclusion This study shows that silver nanoparticles with sizes in the 6-8 nm range may be grown inside a mesoporous titania films by using ammonium silver salt impregnation and chemical reductive treatment. A second population of larger silver nanocrystals (size between 20 and 120 nm) is also formed on the film surface by diffusion and aggregation. Differences between the two reductive treatments are found because it is shown by RBS analysis that NaBH4 treatment induces the silver nanoparticle accumulation in the subsurface of the layer, whereas a formaldehyde process induces the formation of silver nanoparticles almost within the whole thickness of the titania film. Moreover, irradiation with visible light causes the reversible photooxidation of silver nanoparticles by titania, which has been clearly related to a strong decrease of the size of the silver nanoparticles (TEM, SEM, and AFM observations). These Ag-TiO2 nanocomposites described in this work are expected to find important applications for the fabrication of photochromic materials, optical filters, or sensors. Acknowledgment. This work is supported by the ANR POMESCO project. We sincerely acknowledge Pr. M. Roman (Laboratoire des Sciences Analytiques, UMR CNRS 5180, B^at Jules Raulin, Universit
e Claude Bernard-Lyon 1, 43 Bd 11 novembre 1918, 69622 Villeurbanne, France) for fruitful discussions and the reviewers for their comments. We thank C. Grossiard (Science et Surface, Ecully) for performing XPS analysis. Y.B. acknowledges the R
egion Rh^one Alpes for financial support.

Langmuir 2010, 26(2), 1199–1206