Photoluminescent Silica Nanotubes and Nanodisks Prepared by the

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

Photoluminescent Silica Nanotubes and Nanodisks Prepared by the Reverse Micelle Sol-Gel Method Subhasree Banerjee and Anindya Datta* Department of Chemistry, Indian Institute of Technology Bombay, Powai, Mumbai 400 076, India Received June 24, 2009. Revised Manuscript Received August 26, 2009 The reverse micelle sol-gel method was used earlier to prepare silica nanotubes, in aerosol OT/ n-heptane/water microemulsions containing FeCl3. The present communication reports the remarkable effect of the amount of water in the microemulsions on the shape, size, and spectral properties of the silica nanostructures formed. Nanotubes are formed, as expected, at lower water contents. However, for higher water contents, nanodisks form in predominance. This rather surprising observation indicates the formation of flat, disklike water pools in this medium. Notably, a phase separation occurs at higher water contents, and this appears to be essential for the formation of the disklike nanostructures. Hence, we propose that flat water pools form at the interface of the two liquid phases. The nanotubes and nanodisks exhibit blue photoluminescence. The photoluminescence of the nanotubes is more susceptible to quenching by moisture than that of the nanodisks. Luminescence is restored by heating or purging nitrogen or oxygen. Time-resolved photoluminescence studies conform to a model in which the luminescence is ascribed to a particular kind of defect center, with some contribution from surface-associated defects.

Introduction Silica nanomaterials like nanotubes,1 nanorods,2 nanoparticles,3 and nanowires4 have attracted significant interest by virtue of their tunable shape, size, and morphology.5 These nanomaterials exihibit a range of optical,6 electrical,7 and mechanical properties,8 which are often related to their sizes and structures. The Si-OH moieties on the surface of these nanomaterials make them functionalizable using simple silane chemistry.9 Such surface-modified nanomaterials can be useful in catalysis,10 separation,10 detection,11 biolabeling,12 and biomolecule delivery.13 The different approaches for the fabrication of silica nanomaterials include the sol-gel template method,14 biomimicking growth *To whom correspondence should be addressed. Tel: +91 22 2576 7149. Fax: +91 22 2572 3480. E-mail: [email protected].

(1) Jung, J. H.; Yoshida, K.; Shimizu, T. Langmuir 2002, 18, 8724. (2) Zhu, G.; Zou, X. P.; Cheng, J.; Wang, M. F.; Su, Y. Adv. Mater. Res. 2008, 47-50, 367. (3) Jin, Y.; Lohstreter, S.; Pierce, D. T.; Parisien, J.; Wu, M.; Hall, C., III; Zhao, J. X. Chem. Mater. 2008, 20, 4411. (4) Jiang, Z.; Xie, T.; Yuan, X. Y.; Geng, B. Y.; Wu, G. S.; Wang, G. Z.; Meng, J. W.; Zhang, L. D. Appl. Phys. A: Mater. Sci. Process. 2005, 81, 477. (5) Jin, L.; Wang, J.; Cao, G.; Choy, W. C.H. Phys. Lett. A 2008, 372, 4622. (6) Lin, J.; Huang, Y.; Zhang, J.; Gao, J.; Ding, X.; Huang, Z.; Tang, C.; Hu, L.; Chen, D. Chem. Mater. 2007, 19, 2585. (7) Petrovi
c, Z. S.; Javni, I.; Waddon, A.; B
anhegyi, G. J. Appl. Polym. Sci. 2000, 76, 133. (8) Wang, Z. L.; Dai, Z. R.; Gao, R. P.; Bai, Z. G.; Gole, J. L. Appl. Phys. Lett. 2000, 77, 3349. (9) Han, W. S.; Kang, Y.; Lee, S. J.; Lee, H.; Do, Y.; Lee, Y.-A.; Jung, J. H. J. Phys. Chem. B 2005, 109, 20661. (10) Mitchell, D. T.; Lee, S. B.; Trofin, L.; Li, N.; Nevanen, T. K.; S€oderlund, H.; Martin, C. R. J. Am. Chem. Soc. 2002, 124, 11864. (11) Bruchez, M., Jr.; Moronne, M.; Gin, P.; Weiss, S.; Alivisatos, A. P. Science 1998, 281, 2013. (12) Chan, W. C. W.; Nie, S. Science 1998, 281, 2016. (13) Chen, C.-C.; Liu, Y.-C.; Wu, C.-H.; Yeh, C.-C.; Su, M.-T.; Wu, Y.-C. Adv. Mater. 2005, 17, 404. (14) Jung, J. H.; Ono, Y.; Shinkai, S. Langmuir 2000, 16, 1643. (15) Gautier, C.; Lopez, P. J.; Hemadi, M.; Livage, J.; Coradin, T. Langmuir 2006, 22, 9092. (16) M€adler, L.; Kammler, H. K.; Mueller, R.; Pratsinis, S. E. Aerosol Sci. 2002, 33, 369. (17) Yao, N.; Xiong, G.; Yeung, K. L.; Sheng, S.; He, M.; Yang, W.; Liu, X.; Bao, X. Langmuir 2002, 18, 4111.

1172 DOI: 10.1021/la902265e

strategies,15 flame spray pyrolysis,16 ultrasonic synthesis,17 thermal oxidation,18 and chemical vapor deposition.19 The morphology of the nanostructure is often governed by the method of preparation. In addition to the surface-related chemistry, the luminescent nature of these amorphous nanostructures has attributed them with potential applications in optoelectronic devices and optical sensors.9 Photoluminescence (PL) from amorphous silica consists of bands in blue, red, green, and UV regions. Several kinds of defect centers are responsible for these luminescence bands. These defects are intrinsic or extrinsic in nature.20 The origin of the luminescence has been variously ascribed to chemisorbed Si-OH moieties on the outer and inner surfaces of the nanotubes,21 nonbridging oxygen hole centers (NBOHC), tSi-O• in silica matrix,22 oxygen-deficient centers (ODCs),19 and self-trapped excitons (STEs).19 The hydroxyl radicals, tSi-OH, the peroxy linkages, tSi-O-O-Sit, in oxygen-rich species, and irradiation-inducing strained silicon oxygen (tSiO-Sit) bonds are the three precursors of NBOHCs. Neutral oxygen vacancies, tSi-Sit, and 2-fold coordinated silicon defects, tSi-O-Si-O-Sit, are the origins of the ODCs. STEs consist of a peroxy linkage and an E0 center. Aging and heating affect the PL intensity of SiO2 significantly.23 Heating at ∼200-400 °C causes an enhancement of PL intensity, but heating at more than ∼600 °C, postheat treatment at vacumm, causes a reduction in the PL intensity, as does aging. Annealation in an oxygen environment augments the PL intensity, while annealation in a nitrogen environment has an opposite effect.19,24 (18) Hu, J. Q.; Jiang, Y.; Meng, X. M.; Lee, C. S.; Lee, S. T. Chem. Phys. Lett. 2003, 367, 339. (19) Shang, N. G.; Vetter, U.; Gerhards, I.; Hofs€ass, H.; Ronning, C.; Seibt, M. Nanotechnology 2006, 17, 3215. (20) Griscom, D. L.; Friebele, E. J. Phys. Rev. B 1986, 34, 7524. (21) Zhang, M.; Ciocan, E.; Bando, Y.; Wada, K.; Cheng, L. L.; Pirouz, P. Appl. Phys. Lett. 2002, 80, 491. (22) Botti, S.; Coppola, R.; Gourbilleau, F.; Rizk, R. J. Appl. Phys. 2000, 88, 3396. (23) Aboshi, A.; Kurumoto, N.; Yamada, T.; Uchino, T. J. Phys. Chem. C 2007, 111, 8483. (24) Uchino, T.; Kurumoto, N.; Sagawa, N. Phys. Rev. B 2006, 73, 233203.

Published on Web 09/16/2009

Langmuir 2010, 26(2), 1172–1176

Banerjee and Datta

There is a lot of ambiguity about the origin of the blue band, in which we are interested. The first model that seeks to address the issue involves the surface silanol (Si-OH) groups.25 The blue PL intensity is decreased significantly in moisture and upon aging.26 The increase in PL intensity upon heating from 100 to 400 °C is ascribed to the formation of defect centers, by the condensation of silanol groups.24

Article Scheme 1. Mechanism of Formation of (A) the Defect Pair Responsible for Blue PL in Amorphous Silica and (B) the Nonemissive EdgeSharing Dimer in Amorphous Silica, as Proposed by Uchino and Coworkers in Ref 24

2Si-OH f Si-O-Si þ H2 O However, the quenching of PL upon heating beyond 400 °C is not explained by the mechanism. Moreover, the silanol group absorps at about ∼170 nm, not at the 330-360 nm range, excitation at which results in blue PL. A second model proposes carbon impurities as the origin of PL. Nevertheless, this model cannot explain the origin of blue PL from the silica gels that are prepared by the normal hydrolysis of silicon alkoxides using inorganic acids as a catalyst without the involvement of any carbonaceous species.27 Very recently, Uchino and co-workers have proposed an interesting model, in which two geminal silanol groups cocondense to yield a metastable defect pair consisting of dioxasilirane, dSi(O2), and silylene, dSi:, centers (Scheme 1). Density functional theory (DFT) calculations have revealed that the PL excitation wavelength of this defect pair is in the range 240-364 nm. This is in good agrement with the observation that the PL excitation wavelength of the silica gel is in the region ∼350 nm.24 In this model, the PL quenching at higher temperature can be explained by the formation of energetically stable but nonemissive edge and corner sharing SiO4 tetrahedral structures.28 In the experiments of Uchino and co-workers, the PL decay dynamics are found to follow a biexponential decay law.27 The shorter component is ascribed to the electron-hole recombination process, while the long component is attributed to surface-related trapping and detrapping.27,24 Such surface-related defect models have been proposed by Tsybeskov et al. as well.29 With this background, we set out with two major objectives. The first is to investigate the effect of the water content of the microemulsion on the shape and size of the nanomaterials prepared therefrom. The second is to probe the origin of PL of silica nanostructures using steady state and time-resolved spectroscopic techniques. To achieve these objectives, synthesis of silica nanostructures is performed using the reverse micelle-mediated sol-gel (RMSG) method at different w0 (ratio of the number of moles of water to the number of moles of the surfactant values to the reverse micellar structures containing the ionic solution in its core) values.30 The effect of heating and purging of nitrogen and oxygen on the PL has been studied by means of steady state and timeresolved emission spectroscopic techniques.

Materials and Methods Silica nanostructures were synthesized by using the scheme proposed by Jang et al.,30 with the modification that varying amounts of water were used, in an attempt to regulate the shape and size of the silica nanostructures. Sodium dioctyl sulfosuccinate (AOT, Aldrich 98%) was used for the preparation of a (25) Tamura, H.; R€uckschloss, M.; Wirschem, T.; Veprek, S. Appl. Phys. Lett. 1994, 65, 1537. (26) Loni, A.; Simons, A. J.; Calcott, P. D. J.; Canham, L. T. J. Appl. Phys. 1995, 77, 3557. (27) Nakazaki, Y.; Fujita, K.; Tanaka, K.; Uchino, T. J. Phys. Chem. C 2008, 112, 10878. (28) Kurumoto, N.; Yamada, T.; Uchino, T. J. Non-Cryst. Solid 2007, 353, 684. (29) Tsybeskov, L.; Vandyshev, J. V.; Fauchet, P. M. Phys. Rev. B 1994, 49, 7821. (30) Jang, J.; Yoon, H. Adv. Mater. 2004, 16, 799.

Langmuir 2010, 26(2), 1172–1176

cylindrical template. It was purified, dried,31,32 and dissolved in freshly distilled n-heptane [high-performance liquid chromatography (HPLC) grade, Spectrochem, Mumbai, India]. Saturated solutions of FeCl3 (98%, AR grade from Merck) in Millipore water were added to this solution in varying quantities. Tetraethyl orthosilicate (TEOS, g99%, Aldrich) was used as the silica precursor. NaOH (AR grade) from Merck was used to cocondense the silica gel. The ionic salts were removed from the products by soaking in HCl solution overnight. Finally, the product was washed repeatedly with water and ethanol and dried in an oven at 100 °C. The nanostructures formed by the RMSG method were imaged by a transmission electron microscope (TEM) (Philips CM 200), scanning electron microscope (SEM, HITACHI, S-3400N), and atomic force microscope (Nanoscope IV, Veeco). Twenty milligrams of the silica nanostructures was dispersed in 15 mL of acetone and sonicated for 15 min. A drop of this dispersion was taken on metallic stub. After the solvent evaporated, the sample was coated with gold, and then, the SEM micrograph was obtained. Energy dispersive X-ray spectroscopy (EDX) was performed on the same SEM instrument using Thermo Noran Ultra dry detector with Peltier cooling. These nanomaterials were also characterized by using a PerkinElmer Spectrum One FT-IR spectrometer, SII Diamond PerkinElmer thermogravimetric differential analyzer (TG-DTA), and Philips X’pert X-ray diffraction system. The steady state absorption spectra were recorded on a JASCO V530 spectrophotometer, and steady state PL spectra were recorded on a Varian Cary with an Eclipse fluorescence spectrophotometer λex = 350 nm, with a bandwidth of 5 nm. The PL spectra were recorded from 25 mg per mL dispersions of the nanotubes/nanodisks in HPLC grade ethanol from Spectrochem. Time-resolved PL decays were recorded in a time-correlated single photon counting (TCSPC) system, from IBH (United Kingdom), with λex = 341 nm. The full width at half-maximum of the instrument response function was 800 ps. The PL decays were collected with an emission polarizer at a magic angle 54.7° and were analyzed by using IBH DAS 8.2 software.33,34

Results and Discussion Upon increasing the water content in the microemulsions, the nature of the mixture changes markedly. A yellow viscous gel is formed for w0 =2 and 3. A yellow transparent solution is formed at w0 = 4 and 5. For w0 g 5, the solution separates into two distinct phases. Upon adding TEOS to these solutions, silica nanostructures are formed. The amorphous structure of the silica nanotubes and nanodisks is confirmed from the X-ray diffraction pattern (Figure S1 of the Supporting Information). The IR spectra (Figure S2 of the Supporting Information) contain the C-H bond frequency of organic molecules (1073 nm) and the (31) Das, S.; Datta, A.; Bhattacharyya, K. J. Phys. Chem. 1997, A101, 3299. (32) Sarkar, N.; Datta, A.; Das, S.; Bhattacharyya, K. J. Phys. Chem. 1996, 100, 15483. (33) Patel, S.; Datta, A. J. Phys. Chem. B 2007, 111, 10557. (34) Panda, D.; Khatua, S.; Datta, A. J. Phys. Chem. B 2007, 111, 1648.

DOI: 10.1021/la902265e

1173

Article

Banerjee and Datta

Figure 1. Effect of water content of the microemulsions on the shape and size of the silica nanostructures formed therefrom: Transmission electron micrographs of silica nanotubes prepared in microemulsions with (A) w0 =2 and (B) w0 =4. Scanning electron micrographs of silica nanodisks prepared in microemulsions with (C) w0 = 10 and (D) w0 = 22. The microemulsions contaned 17 M FeCl3.

O-H bonds (3424 nm). The contribution from the C-H bond is due to the organic molecules, indicating the formation of silica nanomaterials in the water-surfactant template interface. The O-H stretching band indicates the existence of the surface silanol groups. Interestingly, not only the size but also the shape of the nanostructures, is governed dramatically by the water content (Figure 1). Nanotubes are formed from the solutions with w0 =2, 3, and 4, as is expected. The diameters of the tubes vary from 125 to 363 nm. The distribution of diameters is narrower at w0 = 4 (Figure 1 B). However, as w0 values are increased further, flat disklike structures are found to be formed. The relative amount of disks increases with an increase in the w0 value. The diameters of the nanodisks prepared at w0 = 10 are 730 ( 15 nm (Figure 1C). For w0 =22, the diameter is about 1 μm (Figure 1D). For w0 =40, two size distributions are obtained, centered at 2 μm and 350 ( 25 nm (Figure S4A,B of the Supporting Information). The thickness of the nanodisks is 1-7 nm (Figure S3 of the Supporting Information). The narrow size distribution of the nanotube is rather remarkable. More remarkable, perhaps, is the formation of flat disks in the RMSG method, as the shape of the nanostructures is governed by the shape of the water pool in the microemulsion. 1174 DOI: 10.1021/la902265e

The formation of the nanodisks indicates flat water pools over high water content. This is surprising, as one would expect the water pools to be three-dimensional. Notably, these nanodisks are formed only from solutions that undergo phase separation. So, it appears that a flat water pool occurs at the interface of the two layers. Whereas it is known that water pools of many interesting shapes can form in microemulsions, depending upon water content,35 the occurrence of uniform, flat water pools has not been reported in the literature, to the best of our knowledge. The really small values of the thickness of the silica disks formed from these media further support our contention that they are formed from flat, circular disk-shaped water pools at the interface of the two immiscible liquid phases. This is reminiscent of the formation of thin films of silica at oil-water interface, reported recently.36 It may be mentioned here that the microscopic morphology of nanostructures may depend on various factors like the water content, pH, and gelation time.37 In the SEM image, the nanodisks seem to be aggregated (Figure 1C,D). Hence, it may be (35) Pileni, M. P. Nat. Mater. 2003, 2, 145. (36) Kulkarni, M. M.; Bandyopadhyaya, R.; Sharma, A. J. Mater. Chem. 2008, 18, 1021. (37) Hench, L. L.; West, J. K. Chem. Rev. 1990, 90, 33.

Langmuir 2010, 26(2), 1172–1176

Banerjee and Datta

Article Table 1. PL Lifetimes and Their Amplitudes of Silica Nanostructures Prepared at w0 = 2 and 22 Conditionsa λem (nm)

A1

τ1 (ns)

A2

τ2 (ns)

χ2

410 430 460

0.78 1.24 0.22 2.51 1.04 0.88 1.37 0.12 3.24 1.03 0.88 1.38 0.12 3.47 1.02 a The decays are fitted with a biexponential function: I(t) = I(0)[A1 exp(-t/τ1) þ A2 exp(-t/τ2)], where τ1 and τ2 are the two lifetimes and A1 and A2 are their amplitudes. A1 þ A2 = 1; λex = 341 nm.

Figure 2. Enhancement of PL intensity as a result of (A) nitrogen purging and (B) oxygen purging of the ethanolic dispersion of SNT prepared at w0 = 2.

argued that the “nanodisks” do not form due to the existence of water pools but, rather, arise out of the existence of water pools of irregular shape. However, this argument does not seem to be viable due to two reasons: An irregular water pool would yield grotesque nanostructures. It is highly unlikely that one could get “aggregated disks” of such uniform size and shape from such water pools. Besides, the disks get deaggregated upon dilution of the acetone solution used to prepare the sample for the SEM experiment (Figure S4C of the Supporting Information). So, it appears that they are individual disks that form from flat, circular water pools at the interface of the two liquid layers. The aggregation is merely due to the high concentration used for the SEM experiment. From thermogravimetric analysis (Figure S5 of the Supporting Information) of nanotubes, 10-12% weight loss is observed in the temperature range between 32 and 120 °C. This weight loss is likely to be due to the loss of water molecules adsorbed physically on the surface of the nanotubes. The second weight loss of about 3-6% occurs in the temperature range 110-400 °C. This is ascribed to the cocondensation of Si-OH groups with elimination of water molecules. For the disks, 10% weight loss occurs in the range 33-130 °C, while in the second stage, 2-4% weight loss happens in the range 110-320 °C. The absorption spectra for both the nanotubes and the nanodisks (Figure S6A of the Supporting Information) dispersed in ethanol are featureless, by and large, with a shoulder at ∼350 nm. The PL spectrum of the silica nanotube dispersed in ethanol, when excited at 350 nm, matches the spectra reported by Jang et al. (Figure S7 of the Supporting Information).30 The PL is quenched upon aging for a couple of days but is restored upon heating to 100 °C. Nitrogen purging causes a similar restoration of PL intensity (Figure 2A). The PL excitation spectrum exhibits a maximum at 350 nm (Figure S6B-E of the Supporting Information). This is in agreement with the earlier observations with amorphous SiO2 and with the theoretical study of Uchino and co-workers, which identifies the defect pair responsible for the PL. The quenching of PL intensity upon aging is rationalized by the coverage of the surface with Si-OH or physisorbed H2O.21 Upon heating, the Si-OH groups are expected to cocondense, Langmuir 2010, 26(2), 1172–1176

and the physisorbed water or impurities are expected to get removed. Purging with nitrogen is expected to have the same effect. This is the reason why the PL intensity is restored upon heating or nitrogen purging. Bubbling of oxygen gas through the dispersion of nanotubes and nanodisks causes the PL intensity to increase significantly (Figure 2B). A marked change in spectral shape occurs concomitantly, indicating the generation of more defect centers. This might look surprising per se, as oxygen is known as a good fluorescence quencher. The enhancement of PL with molecular oxygen indicates the occurrence of oxidation. The most potent candidates for oxidation are the ODCs. In silica nanomaterials, ODCs are neutral oxygen vacancies, tSi-Sit, 2-fold coordinated silicon defects, tSi-O-Si-O-Sit, and STEs.19 The STE is an electron-hole pair, consisting of an E0 center, tSi •, and peroxy linkage, tSi-O-O-Sit. All of these defect centers are intrinsic or extrinsic.20 Intrinsic defects are exclusively incorporated during growth and remain localized, for example, radiolytic displacements of oxygen from an otherwise perfect Si-O-Si network to produce intrinsic defects. Si-O-S f Si•þ Si þ O þ eþ An extrinsic process, on the other hand, involves the conversion of precursor defects to other defects, for example, the generation of E0 centers. Si-Si f Si•þ Si þ eþ Si-H f Si• þ • H All of the defects, intrinsic or extrinsic, are highly sensitive to environment effects, for example, heat, radiation, gas flow, and chemical reagents. Upon oxygen purging, neutral oxygen vacancies, tSi-Sit gets oxidized to form peroxy linkage, tSiO-O-Sit. This peroxy linkage is very energetically strained. So, it relaxes to form E0 (tSi•), leading to the increase of PL intensity. Apart from this, there is also a probability of the formation of another kind of defect center. Upon oxygen purging, molecular oxygen gets diffused into the interstitial voids of amorphous silica. After that, this molecular oxygen reacts with E0 (tSi•) centers to form superoxide radicals. Si• þ O2 f Si-O-O• Thus, the augmentation in the PL intensity, upon purging oxygen, may be explained by the formation of new defect centers, as has been discussed above. A potential problem with such an explanation is that peroxy and superoxide radicals are believed to be nonemissive in silica matrix.38 Besides, the enhancement of PL occurs with nitrogen as well as oxygen. So, it is also possible that (38) Skuja, L. J. Non-Cryst. Solids 1998, 239, 16.

DOI: 10.1021/la902265e

1175

Article

Banerjee and Datta

Table 2. Comparison of PL Lifetimes and Their Amplitudes of Silica Nanotubes and Nanodisks (SNT and SND, Respectively) Prepared from Reverse Micelles with w0 = 2 and 22, Respectivelya SNT (w0 = 2) sample

A1

N2-purged (λem = 460 nm) O2-purged (λem = 460 nm) a λex = 341 nm.

0.88 0.84

τ1 (ns) 1.38 1.33

A2 0.12 0.16

SND (w0 = 22) τ2 (ns) 3.47 2.46

the enhancement in PL is due to the change in microscopic morphology of the nanostructures, as a result of purging of the gases. Time-resolved PL is recorded with λex = 341 nm to probe the origin of the PL further. Decays are recorded at three λem values in ethanolic suspension of silica nanostructures, prepared from solutions with w0 = 2, 3, 4, 10, 22, and 40. They are found to be biexponential, as observed in silica gels.27 Lifetimes at the different emission wavelengths, corresponding to the three different PL bands, are only slightly different from each other. The short component is almost identical for all w0 values. For the long components, very slight differences (2.5-3.5 ns) are present (Figures S8 and S9 of the Supporting Information and Table 1). From the time-resolved PL studies at different λem values, it is apparent that the PL at three bands in the blue region is not likely to have different origins, unlike what has been proposed in some earlier studies.39 Rather, the origin is likely to be the same, with some contribution from a surface-related trapping-detrapping process, as has been observed in silica gels. A comparison of the PL decays of nitrogen-purged nanotube and nanodisk dispersions (Figure S10A of the Supporting Information) reveals that the contribution of the longer component is larger in the nanotube than in the nanodisk. This result is in line with the Uchino model, in which the long component is really a pointer for deviation from single exponential behavior and whose contribution increases with an increase in the porosity of the silica gel.27 The surface area of the nanotubes is much more than nanodisks. In other words, the nanotubes are more porous than the nanodisks. Hence, the surface-related defect centers are larger in number in the case of nanotubes, as is reflected in the larger contribution of a long lifetime component in the case of nanotubes. Upon purging nitrogen as well as oxygen, both the nanotubes and the nanodisks become more luminescent, as has been discussed already (Figure 2). However, the lifetimes do not increase to any significant extent (Figure S10 of the Supporting Information and Table 1). This indicates that the increase in the intensity is mostly due to an increase in the number of defect centers. Moreover, a difference in the long component of the PL decays of the nanotubes and nanodisks is observed (Figure S10 of the Supporting Information). The later parts of the decay are slower in the nanotube for the nitrogen-purged samples, while they are faster for oxygen-purged samples (Figure S10A,B of the Supporting Information). In other words, the contribution of the longer component is greater in the nitrogen-purged samples than in oxygen-purged samples (Table 2). Its greater contribution in the oxygen-purged samples is in line with the contention of the generation of new defect centers by reaction with oxygen, as has been proposed in the earlier section. Thus, the results of the (39) Kar, S.; Chaudhuri, S. Solid State Commun. 2005, 133, 151.

1176 DOI: 10.1021/la902265e

χ2

A1

1.02 1.05

0.83 0.97

τ1 (ns) 1.38 1.46

A2 0.17 0.03

τ2 (ns) 2.28 5.22

χ2 1.02 1.06

time-resolved PL experiments are also in line with the Uchino model.

Conclusions Reverse micellar sol-gel synthesis of silica nanostructures is performed using AOT reverse micelle as the nanoreactor. It has been found that the shape of these nanomaterials is mainly governed by the shape of the reverse micellar template. At a lower water content, the shape of the silica nanostructure formed is tubular. At higher water contents, however, nanodisks are formed. These structures are likely to have applications in nanophotonics, optical biosensing, and metamaterials research40,41 by virtue of their two-dimensional, thin, platelike structures. The formation of disks indicates the presence of a flat water pool at the interface of the two liquid phases. This point will be explored in the future. The biexponential decay of PL of these nanostructures is in line with the Uchino model, where the short component is attributed to an electron-hole recombination process and the long component is attributed to a surface-related trapping and detrapping process, which is affected by the change in the shape of the surface. The blue PL is very sensitive by the environmental effect and is enhanced by annealing and nitrogen purging. This enhancement is mainly due to the increase in the number of defect centers. The effect of surface porosity is demonstrated by the difference in the longer lifetimes of the nanotubes and the nanodisks. New defect centers are generated by oxygen purging as well. To summarize, silica nanodisks have been prepared for the first time, using the RMSG method. The origin of the PL in them, as well as silica nanotubes, seems to arise from similar local defect centers, as has been proposed by Uchino and co-workers. Acknowledgment. This work is supported by CSIR project number 01(2277)/08/EMR-II. S.B. thanks CSIR for a Senior Research Fellowship. The SEM images were recorded at the Department of Metallurgy and Materials Science, IIT Bombay, with the support and guidance of Dr. S. L. Kamath. The helpful suggestions from Prof. Indradev Samajdar and Mr. E. Siva Subramaniam Iyer are acknowledged. We thank Prof. S. P. Moulik and A. Layek for useful suggestions. The TEM images were recorded at SAIF, IIT Bombay. The AFM data were recorded at the SPM Facility, Department of Physics, IIT Bombay. Supporting Information Available: X-ray diffraction spectra, IR spectra, TGA, DTA data, additional TEM and SEM micrographs, AFM data, additional PL spectra, decays, and table of PL lifetimes. This material is available free of charge via the Internet at http://pubs.acs.org. (40) Dmitriev, A.; Pakizeh, T.; K€all, M.; Sutherland, D. S. Small 2007, 3, 294. (41) Qin, L.; Banholzer, M. J.; Millstone, J. E.; Mirkin, C. A. Nano Lett. 2007, 7, 3849.

Langmuir 2010, 26(2), 1172–1176