Gold nanoparticles incorporated mesoporous silica thin films of varied gold contents and their well-tuned third-order optical nonlinearities
Qiang Lu,Fangming Cui,Changqing Dong,Zile Hua,Jianlin Shi
DOI: https://doi.org/10.1016/j.optmat.2011.02.024
IF: 3.754
2011-01-01
Optical Materials
Abstract:Highlights ► Gold nanoparticles incorporated mesoporous silica thin films of varied gold contents from 4.64 to 29.15 wt.% were synthesized. ► The films showed increased third-order optical nonlinear susceptibility χ (3) from 4.26 × 10 −11 to 9.24 × 10 −10 esu at increased gold contents. ► The dependence of χ (3) on gold content can be described by an exponent function when the gold contents were below 30 wt.%. Abstract Gold nanoparticles (NPs) incorporated mesoporous silica thin films (MSTFs) of varied gold contents from 4.64 to 29.15 wt.% were synthesized through a refined chemical modification to the mesopore surface using different amounts of silane with amino end group. The microstructures of the composite thin films were characterized and the off-resonant third-order optical nonlinearities of the composite thin films were investigated by Z-scan technique at 1064 nm. The resultant composite thin films showed increased third-order optical nonlinear susceptibility ( χ (3) ) from 4.26 × 10 −11 to 9.24 × 10 −10 esu at increased gold contents. The dependence of χ (3) on gold content have been discussed, which can be described by an exponent function y = y 0 + Ae − x / t when the gold contents of the composite thin films were below 30 wt.%. Keywords Gold nanoparticle Mesoporous silica thin films Chemical modification Third-order optical nonlinearity 1 Introduction The noble metal nanoparticles (NPs) incorporated dielectric matrix solid thin films with outstanding third-order optical nonlinearities have been widely investigated due to their potential application prospect in the next generation of optical communication and logical processing devices [1–5] . Among them, the noble metal NPs incorporated solid thin films which were prepared in early days by physical method, such as co-sputtering, suffered from the drawbacks of uncontrollable particle size and spatial distributions, and consequently the nonlinear optical properties might be significantly affected. Recently the noble metal NPs incorporated mesoporous thin films were developed which successfully conquered these problems and showed a relatively high third-order nonlinear optical susceptibility ( χ (3) ). For example, the χ (3) value of gold NPs incorporated mesoporous silica thin films (MSTFs) with a 16.2 wt.% gold content was approximately 5.62 × 10 −11 esu [5] . However, due to the metal content in the mesoporous composite thin films could not be well-tuned owing to the difficulties in the metal NP incorporation with high amounts, the dependence of the third-order optical nonlinearities on the noble metal contents of the mesoporous composite thin films were rarely reported except for only one example of gold NPs incorporated mesoporous titania thin films [6] . Although the noble metal NPs have been introduced into the mesoporous channels by a variety of methods, such as wet impregnation under reduced pressure, ion exchanging and chemical surface modification [5–8] , the noble metal NPs incorporated mesoporous materials (especially mesoporous thin films) with different metal contents have not been paid enough attentions due to the difficulties in synthesis, therefore it is still a challenge to obtain varied noble metal NP content in MSTFs under a controlled manner. Meanwhile, the third-order optical nonlinearity of the noble metal NPs incorporated composite material has been suggested to be highly dependent on the noble metal content of the composite material [9–11] . In order to investigate the dependence of χ (3) on the gold content of the thin films and find out the optimal composition of the gold NPs incorporated MSTFs with as high as possible of χ (3) , it is necessary to synthesize gold NPs incorporated MSTFs with varied gold contents and investigate the dependence of third-order optical nonlinearities of the materials on the gold contents. In this paper, we have succeeded in tuning the gold contents from 4.64 to 29.15 wt.% in the gold NPs incorporated MSTFs by a refined surface modification method (through carefully modifying the mesopore surface with different amounts of silane with amino end groups, which could subsequently absorb different amounts of gold precursor into the mesoporous channels). The third-order optical nonlinearities of the gold NPs incorporated MSTFs were examined by a simple Z-scan technique at an off-resonant wavelength 1064 nm. And the gold NPs incorporated MSTFs with different gold contents presented remarkably increased χ (3) values from 4.26 × 10 −11 to 9.24 × 10 −10 esu with the increase of the gold contents. The dependence of χ (3) on gold contents can be described by an exponent function y = y 0 + Ae − x / t when the gold contents of the composite thin films were below 30 wt.%. 2 Experimental 2.1 Preparations MSTFs were prepared according to the reported method [12] . The thin films were then air-dried for 24 h followed by drying at 100 °C for 12 h. Surfactant templates were removed by ethanol refluxing at 80 °C for 24 h. The template-removed MSTFs were dried in vacuum oven for 12 h. Then, dried MSTFs (coated on both surface of a glass substrate with 25 × 50 mm size) were immersed into 80 mL anhydrous toluene with different amount of amino silane (N-(2-aminoethyl)-3-aminopropyltrimethoxysilane) and refluxed for 3 h at 120 °C. After that, the amino silane modified MSTFs were rinsed with anhydrous toluene and then dried in vacuum oven at 80 °C for 12 h. The dried thin films were immersed into HAuCl 4 aqueous solution and kept for 2 h. Then, the thin films were taken out and rinsed with deionic water. Finally, the gold NPs incorporated MSTFs were obtained by reducing the gold precursor in the mesoporous channels under H 2 /Ar (H 2 : 6 vol.%) atmosphere at 300 °C for 2 h. The gold NPs incorporated MSTFs with different gold contents were labeled as AS-1, AS-2, AS-3 and AS-4. The detailed silane concentration of each toluene solution, the gold contents, the optical nonlinear refractive indexes and third-order optical nonlinear susceptibilities of the samples were listed in Table 1 . 2.2 Characterizations Small-angle X-ray diffraction (SAXRD) was used to characterize the meso-structure of the samples with Rigaku Rotaflex diffractometer equipped with a rotating anode and using Cu Kα radiation. Transmission electron microscopy (TEM), selective area electron diffraction (SAED) and simultaneous energy-dispersive spectroscopy (EDS) were used to analyze the microstructure and the gold contents of the samples using JEOL 2100F electron microscope operating at 200 kV. Scanning electron microscopy (SEM) analysis was measured on JEOL JSM-6700 Field Emission Scanning Electron Microscope. UV–visible absorption spectra were measured on Shimadzu UV-3101. The thicknesses and linear refractive indexes at 1064 nm of the composite thin films were measured by SGC-10 film thickness measurement apparatus. The transmittances of the composite thin films were measured by dividing the transmitted laser energy with the incident laser energy used in Z-scan experiment. Z-scan technique was used to detect the optical nonlinear refractions and absorptions of the samples using a mode-locked Nd:YAG laser: 40 ps duration, 10 Hz frequency and 1064 nm wavelength. The laser came through an aperture of 3 mm in diameter then across a lens with 70 mm focal length and focused to a beam waist of 35 μm at the focus. And the laser beam power density at the focus was 8.502 × 10 10 (±5%) W/cm 2 . In order to use Z-scan technique accurately, the experimental conditions were carefully controlled. First, we used two identical optical detectors and an energy ratio meter to correct the initial data. The influence of the fluctuation from the incident laser power was eliminated by dividing the transmitted power by the power obtained at the reference detector. Second, we set the frequency of laser pulse to 10 Hz to reduce the possible thermal accumulation effect. Finally, the transmission rate of the sample was calculated by averaging the detected data from 500 laser pulses. 3 Results and discussions Fig. 1 shows the SAXRD patterns of the blank MSTF and the gold NPs incorporated MSTFs with different gold contents. The distinct diffraction peak at 2 θ = 1.1° presented in the curve a in Fig. 1 can be indexed as (1 0 0) plane reflection of the two-dimensional hexagonal meso-structure of the as-prepared MSTFs [12] . After the surfactant templates were removed and the gold NPs were incorporated within the composite thin films by calcination at 300 °C for 2 h, the (1 0 0) plane diffraction peak moved towards to a higher 2 θ region due to the further contraction of the mesoporous silica framework, as seen curves b–e in Fig. 1 . And we can also see that the (1 0 0) diffraction peak intensity of the gold NPs incorporated MSTFs decreased from sample AS-1 to AS-4 where it nearly disappeared. This is due to the gradually weakened diffraction contrast between the silica framework and the pore channels resulted from the gold content increasing from sample AS-1 to AS-4, which in turn confirms that the gold NPs have been indeed incorporated into the mesoporous channels with increased gold contents from sample AS-1 to AS-4 [5] . Similar experimental result has been reported in literature [13] , where the higher guest content in the mesoporous composite thin films resulted in the decreased diffraction peak intensity. Fig. 2 presents the TEM images of the blank MSTF and the gold NPs incorporated MSTFs with different gold contents and the SAED pattern of the incorporated gold NPs. The as-prepared blank MSTF shows clear ordered two-dimensional hexagonal structured mesoporous channels, as seen in Fig. 2 a. From Fig. 2 b–e, which present samples AS-1, AS-2, AS-3 and AS-4, we can see that the gold NPs (the black dots in the images) dispersed homogeneously within the MSTFs with an average diameter of about 6 nm due to the size confinement effect of the mesoporous channels [5] . And the gradually increased black dots densities from Fig. 2 b–e indicate the gold contents gradually increased from sample AS-1 to AS-4, which is consistent with the above SAXRD analysis. The characteristic SAED pattern of the gold NPs in Fig. 2 f can be indexed as (1 1 1), (2 0 0), (2 2 0) and (3 1 0) planes of face-centered cubic gold crystals [6] , indicating that the gold NPs were well crystallized. The quantitative values of the gold contents in each sample were measured by simultaneous EDS technique and the results were listed in Table 1 . We obtained each of the value by averaging the different data from five different points on each sample. Fig. 3 shows a typical EDS spectrum of the gold NPs incorporated MSTFs, which confirms the gold existence within the composite thin films. From Table 1 , we can clearly see that the gold contents of the samples increased from 4.64 wt.% in sample AS-1 to 29.15 wt.% in sample AS-4, which was in accordance with the increase of the silane concentration in each toluene solution from 2.9 × 10 −3 to 11.6 × 10 −3 mol/L. This indicates that the gust gold contents of the samples can be effectively tuned by changing the silane concentrations. With the increase of the silane concentrations of the toluene solutions, mesoporous channel surface could be modified more completely and as a result higher amount of gold precursor could be introduced, resulting in the final increased gold contents in the MSTFs. However, a great deal of experiments showed that, the gold contents in MSTFs cannot be enhanced unlimitedly with the increase of the silane concentrations, i.e. there is an optimal and critical silane concentration at which the highest amount of amino groups/gold precursors can be introduced into the mesoporous channels. When the silane concentration was over the critical value ∼11.6 × 10 −3 mol/L in our experiments, the gold content of the sample would decrease rather than increase. This is because, at over high silane concentration, the pore channels may be blocked by the silane molecules and less Au precursors/Au NPs could be finally introduced. In the present case, at 11.6 × 10 −3 mol/L of silane concentration, the highest of gold content of 29.15 wt.% could be achieved. In order to confirm the locations of the gold NPs within the mesoporous thin films, a cross-sectional back scattered SEM image of the gold NPs incorporated MSTFs (sample AS-4) was taken and shown in Fig. 4 . The section indicated by two arrows in Fig. 4 presents the composite thin film coated on the substrate. Due to the much higher atomic number of Au than the SiO 2 background, the white dots in the SEM image represent the gold NPs which were embedded and dispersed homogeneously in MSTFs all through the whole cross-section area of the mesoporous thin films and with a narrow size distribution in diameter. And this is consistent with the above TEM results. The average film thickness of the gold NPs incorporated MSTFs is about 218 nm measured from the cross-sectional SEM image, which is in good agreement with the film thickness (215 nm) measured by optical interference method on a SGC-10 film thickness measurement apparatus. Fig. 5 presents the UV–visible absorption spectra of blank MSTF and gold NPs incorporated MSTFs with different gold contents. The blank MSTF do not show any absorption peak in the region from 400 to 800 nm as seen curve a in Fig. 5 . Whereas all the gold NPs incorporated MSTFs show apparent absorption peaks at about 535 nm as seen from curve b to e in Fig. 5 , which correspond to the surface plasmon resonance (SPR) absorption of gold NPs [14] . The different intensities of the SPR absorption peaks in curves b–e in Fig. 5 were in accordance with the different gold contents in samples AS-1, AS-2, AS-3 and AS-4 with the lowest in sample AS-1 and the highest in sample AS-4, as already indicated by the above XRD and TEM results. The increased background absorption of the gold NPs incorporated MSTFs, compared with that of the blank MSTF, originates from an inter-band transition of valence electrons of the gold NPs from the d band to the Fermi surface [5] . In this report, we mainly focus on the off-resonant optical nonlinearities of the gold NPs incorporated MSTFs, which were examined by Z-scan technique using an incident laser at 1064 nm. Because the resonant excitation of the noble metal NPs would lead to heat accumulation in the materials, which is undesirable for the practical optical device applications. The third-order optical nonlinearities of the materials measured at/around resonant region were usually ten times higher than that measured in off-resonant region due to the local field enhancement by SPR [6] . Z-scan technique is an effective method to determine the optical nonlinear refraction and absorption of the detected materials by recording the dependence of the signal intensity on the sample position. The closed aperture (CA) scheme allowed the determination of both sign and value of the nonlinear index ( n 2 ), and the open aperture (OA) scheme was used for the measurement of the nonlinear absorption coefficient ( β ). When the nonlinear refraction and absorption are simultaneously present in the sample, the normalize transmittance T ( z ) dependence in the CA output can be written as the following [15] : (1) T ( z ) = 1 + 4 x ( x 2 + 9 ) ( x 2 + 1 ) Δ Φ 0 - 2 ( x 2 + 3 ) ( x 2 + 9 ) ( x 2 + 1 ) Δ Ψ 0 where x = z / z 0 , z 0 = kw 0 2 / 2 is the diffraction length, w 0 (35 μm) is the laser beam waist radius at the focal plane at 1/ e 2 intensity level, and k = 2 π / λ is the wave vector magnitude, λ = 1064 nm. Δ Φ 0 = kγI 0 L eff and Δ Ψ 0 = βI 0 L eff /2, where γ is the nonlinear refractive index measured in SI units, γ is connected with n 2 measured in esu units by a following relation n 2 [esu] = ( cn 0 /40 π ) γ [m 2 W −1 ], where c is the light velocity and n 0 (1.42) is the linear refractive index of the sample. I 0 (5.72 × 10 14 W/m 2 ) is the power of laser energy at the focus. L eff = [1 − exp(− αL )]/ α is the effective length of the investigated sample, L (215 nm) is the real thickness of the composite thin films, α = −ln T / L is the linear absorption coefficient of the sample, T is the transmittance of the sample at 1064 nm, as listed in Table 1 . We can obtain n 2 and β by a nonlinear fitting according to the above T ( z ) dependence formula. Then the third-order optical nonlinear susceptibility χ (3) can be considered as a complex value of χ (3) = Re χ (3) + i Im χ (3) . The real part Re χ (3) and imaginary part Im χ (3) of χ (3) can be calculated using n 2 and β according to the following [6] : (2) Re χ ( 3 ) = ( n 0 / 3 π ) n 2 (3) Im χ ( 3 ) = ( ε 0 c 2 n 0 / ω ) β where ε 0 is permittivity of vacuum, and ω is the laser frequency. Fig. 6 a shows a typical CA Z-scan output of the gold NPs incorporated MSTFs. The red solid line is the theoretical fitting (Adjusted R -Square = 0.93668). The profile with a peak followed by a valley in the CA normalized transmittance output represents a self-defocusing nonlinear response of the sample, which was resulted from the optical nonlinear refraction, and a negative nonlinear refractive index of the composite thin films [16,17] . In this study, all the samples did not show nonlinear absorptions probably due to the off-resonance measurement conditions (detected at 1064 nm), as presented by the OA Z-scan output in Fig. 6 b. Therefore, we defined Δ Ψ 0 = 0 during the theoretical fitting. We ultimately obtained n 2 and χ (3) values of the composite thin films through the nonlinear fitting and calculation according to the above equations. The detailed results were listed in Table 1 . Fig. 7 presents the third-order optical susceptibilities of the gold NPs incorporated MSTFs. We can see that the third-order optical nonlinearities of the composite thin films increased by more than 20 times from 4.26 × 10 −11 to 9.24 × 10 −10 esu as the gold contents of the composite thin films increased by about seven times from 4.64 to 29.15 wt.%. This result is consistent with that in literature where the composite thin films were prepared by physical co-sputtering method, when the gold contents were below 30 wt.% [10] . And we tried to fit these data according to an exponent function y = y 0 + Ae −x / t and got an adjusted R -Square value of 0.92648, indicating the experimental data are closed to the fitting result and an acceptable fitting result of the experimental data. The detailed parameters in the exponent function were listed in the inset in Fig 7 . From the fitting curve, we can estimate the χ (3) values of the gold NPs incorporated MSTFs according to the above exponent function when the gold contents of the samples were below 30 wt.%. However, it is uncertain whether the relation between the χ (3) and gold content will still follow the above exponent function at the gold content higher than 30 wt.%, as the gold NPs incorporated MSTFs with gold content higher than 30 wt.% are not available experimentally in the present study. In spite of the experiment limitations, one can still refer that even if significantly higher gold amount could be incorporated in MSTF, the mesoporous structure of MSTF would be further deteriorated to a much larger extent than that in sample AS-4, and gold NPs may aggregate with each other and/or grow much larger due to weakened mesopore confinement effect. And the transmittance of the films would be decreased dramatically, which would worsen the optical nonlinearity of the materials. So an optimal and critical gold content for the highest χ (3) value should be present. 4 Conclusions In conclusion, gold NPs incorporated mesoporous silica thin films of varied gold contents from 4.46 to 29.15 wt.% were successfully synthesized through a refined chemical modification of the silica mesopore surface using different concentrations of silane with amino end group. The resultant composite thin films with different gold contents show well-tuned third-order optical nonlinear susceptibilities, which increased from 4.26 × 10 −11 to 9.24 × 10 −10 esu as the gold contents varied from 4.46 to 29.15 wt.%. An exponent function ( y = y 0 + Ae − x / t ) fitting of the nonlinear optical data was made and the χ (3) value of the gold NPs incorporated MSTFs can be estimated from the exponent function fitting at the gold contents below 30 wt.%. Therefore this report provides us a simple and effective method to tune the third-order optical nonlinearities of the gold NPs incorporated MSTFs by varying the gold contents in MSTFs. Acknowledgments The authors thank the National Science Foundation of China (50976032) and Nature Science Foundation of Beijing (3101001) for financial support. References [1] D. Ricard Ph. Roussignol C. Flytzanis Opt. Lett. 10 1985 511 513 [2] M.J. Bloemer J.W. Haus P.R. Ashley J. Opt. Soc. Am. B 7 1990 790 795 [3] K. Fukumi A. Chayahara K. Kadono T. Sakaguchi Y. Horino M. Miya K. Fujii J. Hayakawa M. Satou J. Appl. Phys. 75 1994 3075 3080 [4] S.T. Selvan T. Hayakava M. Nogami Y. Kobayashi L.M. Liz-Marzan Y. Hamanaka A. Nakamura J. Phys. Chem. B 106 2002 10157 10162 [5] J. Gu J. Shi G. You L. Xiong S. Qian Z. Hua H. Chen Adv. Mater. 17 2005 557 560 [6] F. Cui Z. Hua C. Wei J. Li Z. Gao J. Shi J. Mater. Chem. 19 2009 7632 7637 [7] H. Chen J. Shi Y. Li J. Yan Z. Hua H. Chen D. Yan Adv. Mater. 15 2003 1078 1081 [8] A. Fukuoka H. Araki Y. Sakamoto N. Sugimoto H. Tsukada Y. Kumai Y. Akimoto M. Ichikawa Nano Lett. 2 2002 793 795 [9] H. Liao R. Xiao H. Wang K. Wong G. Wong Appl. Phys. Lett. 72 1998 1817 1820 [10] H. Liao R. Xiao J. Fu P. Yu G. Wong P. Sheng Appl. Phys. Lett. 70 1997 1 3 [11] F. Hache D. Ricard C. Flytzanis J. Opt. Soc. Am. B 3 1986 1647 1655 [12] P. Alberius K. Frindell R. Hayward E. Kramer G. Stucky B. Chmelka Chem. Mater. 14 2002 3284 3294 [13] L. Zhang Z. Hua X. Dong L. Li H. Chen J. Shi J. Mol. Catal. A: Chem. 268 2007 155 162 [14] S. Link M. El-Sayed J. Phys. Chem. B 103 1999 8410 8426 [15] R. Ganeev A. Ryasnyansky A. Stepanov T. Usmanov Phys. Status Solidi b 238 2003 R5 R7 [16] D. Smith Y. Yoon R. Boyd J. Campbell L. Baker R. Crooks M. George J. Appl. Phys. 86 1999 6200 6205 [17] S. Debrus J. Lafait M. May N. Pincon D. Prot C. Sella J. Venturini J. Appl. Phys. 88 2000 4469 4475