Highly dispersed and uniform Fe2O3 nanoparticles (NPs) have been incorporated into the pore channels of SBA-15 mesoporous silica thin films (MSTFs). And such Fe2O3 NPs incorporated MSTFs did not show detectable nonlinear optical (NLO) signals at off-resonance wavelength 1064 nm by Z-scan technique. However after a vacuum heat treatment at 800 °C for 1 h under 6 T magnetic field, the Fe2O3 NPs incorporated MSTFs with very low Fe content (0.8~1.5 at.%) presented distinctive NLO signals with χ(3) value in an order of 10−10 esu. We proposed the physical reason for the NLO property generation to be the magnetic domain orientation of the iron oxide NPs incorporated within the pore channels of the MSTFs by the magnetic field heat treatment.
©2010 Optical Society of America
Noble metal nanoclusters/particles contained dielectric/glass matrixes were investigated widely for the nonlinear optical (NLO) properties [1–4]. However, the metal nanoparticles (NPs) usually tend to grow coarser and agglomerate at elevated temperatures. Alternatively, oxides which possess the similar or even better NLO performances also became very interesting due to their relatively low cost as compared to the noble metals. The 3d transition metal oxides, such as Fe2O3, were preferred due to its inexpensive, highly environmentally stable and environmentally-friendly features as compared to noble metals. In 1995, Ando et al  reported the third-order optical nonlinearities (χ(3)~10−8 esu) of the 3d transition metal oxides (V2O5, Cr2O3, Mn3O4, Fe2O3, Co3O4, CuO and NiO), which were prepared by thermal decomposition at 380 °C in air of metal alkylcarboxylates spin-coated on a glass-plate substrate. For those studies, the light source was a frequency-doubled Q-switched Nd: YAG laser at a wavelength of 532 nm and pulse duration of 7 ns or 35 ps. Other reports [6–8] also demonstrated that the pure iron oxide possessed a relatively high χ(3) value in orders of 10−11 to 10−8 esu measured by the laser with its wavelength at or near the semiconductive iron oxide absorption band (~520 nm) [5,6], such as at 532 nm, 488nm. While the off-resonant (the incident laser wavelength far away from the absorption band, e.g. 1064 nm) third-order NLO properties of iron oxide have been rarely reported so far except for that in literature . With the off-resonant NLO response, one can effectively avoid the heat accumulation in the materials during the laser irradiation comparing to the resonant NLO response which is un-favored for the practical NLO applications of the material. Therefore, the off-resonant NLO property of the material is of great significance for both the research and the application of the NLO materials.
However the reported pure iron oxide thin films [5–7] prepared by sol–gel route still suffered from the uncontrolled particle properties, such as particle size and size distribution. Due to the uniform and ordered mesopore channels ranging from 2~30 nm of the mesoporous materials, these pore channels can be used as micro-container for the guest material incorporation. Moreover, the guest materials confined within the mesoporous channels could be prevented from over-growing and/or aggregation by the confinement of mesoporous channels. Highly dispersed and uniform Fe2O3 NPs have been incorporated into the pore channels of SBA-15 mesoporous silica thin films (MSTFs)  as previously reported by our group, in which the guest NPs were strictly confined inside the mesoporous channels of the host MSTFs due to the molecular level dissolution of the precursor compounds into the hydrophobic core of the surfactant micelles used for mesopore templating. However, the third-order optical nonlinearities of the iron oxide incorporated MSTFs (Fe–MSTFs) have not been reported so far. The preliminary studies in our group showed that such Fe–MSTFs did not show detectable NLO signals in off-resonance regions (1064 nm), probably due to the very low iron oxide content in the composite films compared to the sol–gel derived pure iron oxide thin films.
On the other hand, magnetic field heat treatments have been recognized as an effective method to control the microstructures and modulate the physical properties of the treated materials [11–14]. Therefore, here we synthesized Fe–MSTFs with different iron contents (<1.5 at. %) and treated the Fe–MSTFs under 6 T magnetic field at an elevated temperature (800 °C), hoping for enhanced NLO properties of the materials.
The NLO refractions and absorptions of the prepared materials were measured by Z-scan technique at 1064 nm. We found that the samples which have underwent the magnetic field heat treatment showed distinctive third-order optical nonlinear susceptibility (χ(3)~10−10 esu), while the samples thermally treated at the same temperature but without the magnetic field did not show any nonlinear optical response under the same measuring conditions. We proposed the NLO property generation of the materials by the magnetic field to the magnetic orientation effect of the iron oxides nanoclusters incorporated within MSTFs during the heat treatment under the magnetic field.
Fe–MSTFs were prepared according to literature . Typically, the hydrophobic compound iron (III)-1,1,1-trifluoro-2,4-pentanedione (FeTFA3) was used as iron oxide precursor, then different amounts of FeTFA3 were mixed with the surfactant solution of poly (ethylene-oxide)-poly (propylene oxide)-poly (ethylene-oxide) block copolymer PEO20PPO70PEO20 (Pluronic P123, BASF) and ethanol. The adding of different FeTFA3 quantities gave the different Fe loading contents in the resultant composite thin films. And then the prehydrolyzed silica solution was added into the surfactant solution under stirring. The composite thin films with different Fe loading content, labeled as Fe–MSTF-1, Fe–MSTF-2 and Fe–MSTF-3, were prepared by dip-coating the clean quartz substrate into the precursor solutions with different FeTFA3 concentrations at a withdrawal rate 1 mm s−1 in air. The values of Fe contents can be seen in Table 1 . The films were then air-dried at room temperature for 24 h and followed by calcination at 400 °C for 3 h. The composite thin films changed from colorless to shallow brown appearance after calcination.
The composite thin films (coated on the quartz substrate) were cut into a square of 5 mm by 5 mm in size before magnetic field thermal processing. Then the sample was fixed on a support using a nickel thread with the composite thin film surface parallel to the horizontal while normal to the magnetic field. After the furnace was vacuumed to 10−5 Pa and the magnetic field was loaded to 6 T, the samples began to be heated (ramp: 5 °C min−1) to 800 °C and kept for 1 h at the temperature, then cooled down to room temperature along with the cooling down of the furnace in air. Finally the furnace was de-vacuumed and the magnetic field was unloaded. For comparison, the composite thin films were also heat treated at 800 °C for 1 h under vacuum but without loading magnetic field.
The NLO properties of the samples were detected by Z-scan technique using an incident laser (mode-locked Nd: YAG) at 1064 nm, with 40 ps pulse duration and 10 Hz in frequency, which is far away from the band gap absorption (~224 nm) of the semiconductive iron oxide nanoclusters. The laser intensity was 1.898 × 1011 W/cm2. The laser used in Z-scan experiment is p polarized light. The laser incident beam goes through the films vertical to the film surface. The film thicknesses and linear refractive indexes of the composite thin films were measured to be 251 ( ± 15) nm and 1.42 by a SGC-10 film thickness measurement apparatus using interference method. 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.
A variety of experimental methods, such as VSM and SQUID, were employed and tried to investigate the magnetic properties of the Fe–MSTFs before and after magnetic field heat treatments, unfortunately, there have been great difficulties in obtaining distinctive signals only originated from the Fe–MSTFs due to the extremely low quantity of the samples (~10−3 mg for each piece of sample) treated in the magnetic field. Therefore, we still could not get clear information of the magnetic properties of the samples.
3. Results and Discussions
The mesostructures of the blank MSTF before and Fe-MSTFs after calcination were characterized by small-angle X-ray diffraction (SAXRD) patterns, as presented in Fig. 1 . The diffraction peaks at 2θ = 1.05° and 1.98 o of the MSTF before calcination can be indexed as (100) and (200) reflections of the highly ordered hexagonal (P6mm) mesostructure with (100) plane spacing of about 8.4 nm [15,4]. Compared to the bulk sample, the (110) reflection is missing, suggesting that the films has a highly oriented 2-dimensional hexagonal mesostructure with its pore channels aligned parallel to the substrate plane . After calcination in air at 400 °C for 3 h, the (100) diffraction peak moved towards higher 2θ region due to the further contraction and condensation of the silica framework. The great decrease of the (100) reflection intensity of the calcined samples can also be found. This could be attributed to the formation of the iron oxide nanoclusters inside the mesoporous channels of MSTFs, which decreased the reflection contrast between the silica framework and the pore channels [4,16]. The differences of the (100) reflection intensities among the composite thin films of the different Fe loading amounts imply that the iron oxide nanoclusters have been loaded within the pore channels of the MSTFs.
Figure 2 presents the typical transmission electron microscopic (TEM) images of the composite thin films (sample Fe–MSTF-3) before and after calcination. From the TEM images, we can see that both of the samples show highly ordered hexagonal mesoporous structure with the pore channels parallel to the films surface. And we can also estimate the (100) planes spacing of the composite thin films before and after calcination to be about 8.4 and 7.1 nm, respectively, which is in accordance with the above SAXRD results. Generally, the TEM imaging could allow us to directly observe the profile and distribution of the guest materials incorporated within MSTFs. However, because of the low electron scattering contrast between iron oxide nanoclusters and silica frameworks due to their similar densities, the iron oxide cannot be clearly distinguished from the silica matrix in the TEM image (Fig. 2 (b)). However the lattice spacing contraction after calcination can be seen from the TEM images. The existence and the content of the Fe element of the composite thin films have been verified by the simultaneous energy dispersive spectroscopy (EDS) measurement. The data of the Fe loading content of each sample are listed in Table 1. The EDS spectrum of the composite thin film Fe–MSTF-3 is shown in Fig. 3 .
Figure 4 shows the X-ray photoelectron spectroscopy (XPS) spectrum of the Fe–MSTFs, which present two energy peaks at 715 and 728 eV corresponding to the specific energy levels of Fe 2p3/2 and Fe 2p1/2 confirming the Fe (III) chemical state . Therefore it is clear that it is Fe2O3 NPs incorporated within MSTFs.
The UV-visible absorption spectra of the blank MSTF and the iron oxide incorporated Fe–MSTFs are presented in Fig. 5 . From the spectra, we can see that the blank MSTF did not show any absorption within the detection region (from 200 to 800 nm wavelength), as seen in Fig. 5 (a). While all the iron oxide incorporated samples present narrow absorption bands at around 224 nm, as seen in Fig. 5 (b), (c) and (d), which is a significant blue shift compared with that (at around 520 nm) of the reported sol–gel derived iron oxide thin films [5,6] and can be attributed to broadened energy gap of the semi-conductive Fe2O3 NPs in MSTFs. The gradually increased energy gap absorption intensity from Fig. 5 (b) to (d) is due to the increased Fe loading amounts of the samples. Since the synthesized composite thin films show clear absorption edges in the UV-visible spectra, the direct energy gap ΔEo of the incorporated iron oxide semiconductor NPs can be determined from the intercept of the square of linear absorption coefficient times photon energy (αhν)2 against the incident photon energy (hν) [10,17], as presented in the insets in Fig. 5. The red lines in the insets of Fig. 5 are the linear fittings, which indicate the energy gaps of the samples Fe–MSTF-1, Fe–MSTF-2, and Fe–MSTF-3 to be 4.7, 4.4, and 4.4 eV, respectively, demonstrating a remarkable broadening of the energy band gap as compared to 2.2 eV of the bulk Fe2O3. Such a distinct band gap broadening can be attributed to the quantum size effect of the incorporated iron oxides NPs, as can be understood from the relation between the change of band gap energy and the radius R of the semiconductor NPs: , here Rb is exciton Bohr radius, e is the charge of electron, and ε0 and ε are the vacuum and material dielectric coefficients . From the equation, one can also calculate the size of the iron oxide NPs loaded in sample Fe–MSTF-1 to be about 2.1 nm which is slightly smaller than those (at about 2.2 nm) in samples Fe–MSTF-2 and Fe–MSTF-3, due to the lower Fe content in sample Fe–MSTF-1 than those in Fe–MSTF-2 and Fe–MSTF-3.
The third-order optical nonlinearities of Fe–MSTFs before and after magnetic field heat treatments were measured by Z-scan technique at 1064 nm. With the technique both the sign and magnitudes of the NLO refractive index n2 and the absorption coefficient β of the materials can be determined . When the nonlinear refraction and absorption are simultaneously present, the relation between normalized transmittance T(x) and the sample position z in the closed aperture (CA) output can be written as , here , is the diffraction length, and are related to n2 and β . Through the fitting results by the relation, both n2 and β can be obtained with the possible experimental errors minimized. Then, the third-order optical nonlinear susceptibility χ(3) can be calculated from n2 and β by the method reported in literature .
All the samples with different Fe loading amounts without magnetic field heat treatment did not show any optical nonlinear response demonstrated by Z-scan outputs, so did the samples heat treated without magnetic field. Interestingly, after the heat treatment under the 6 T magnetic field, the sample Fe–MSTF-2 and Fe–MSTF-3 show apparent signals in Z-scan outputs, with a peak followed by a valley in the CA Z-scan normalized transmittance traces, as seen in Fig. 6 (a) , which indicates the negative nonlinear optical refraction index n2 of the samples corresponding to self-defocusing procedure when the sample moved across the laser focus. The solid line in the Fig. 6 (a) is the fitting curve with adjusted R-Square = 0.90817, indicating a good consistence between the experimental results and the theoretical model. And the self-defocusing and negative nonlinear refraction are consistent with that of the reported Fe2O3 NPs prepared by the colloid chemistry method and measured by a Q-switched Nd: YAG laser with 1060 nm wavelength and 15 ns pulse duration .
And the open aperture (OA) Z-scan outputs of the sample Fe–MSTF-2 and Fe–MSTF-3 after magnetic field heat treatment present nonlinear saturation absorption peaks, as indicated in Fig. 6 (b). The saturation absorption presented by OA Z-scan outputs is different from the gold NPs incorporated MSTFs, even with relative high gold content, which did not show any nonlinear absorption under the same measurement condition . The reason may lie in the differences in energy band structures between semiconductor iron oxide NPs and gold NPs and the different interactions of the loaded guests with the silica host. The saturation absorption and negative refraction of the Fe2O3 NPs incorporated MSTFs are very similar with that of the intrazeolite PbS quantum dots measured by Z-scan technique with 1064 nm wavelength and 50 ps pulse width. And the saturation absorption at 1064 nm indicates that the semiconductors behave as optical bleachers , which can be explained by the two-level energy system where the strong laser beam equalizes the population distribution between the two levels everywhere so that there is no and less absorption around the focus, and the beam bleaches a transparent path through the medium. Therefore the OA Z-scan output presents a peak in the normalized transmittance trace.
The third-order optical nonlinear susceptibilities χ(3) of the samples thermally treated under magnetic field were calculated according to the reported equations , and the results were listed in Table 1. From Table 1, we can see that sample Fe–MSTF-1 both before and after magnetic field heat treatment did not show distinguishable NLO responses, probably due to the over-low Fe loading content. Comparatively, for the samples Fe–MSTF-2 and Fe–MSTF-3, only after magnetic field heat treatments, the samples presented significant NLO responses with χ(3) values at a 10−10 esu magnitude order. Although the smaller size of the semiconductor NPs would result in higher NLO signals according to the reports [8,21], the iron oxide NP size difference between the sample Fe–MSTF-1(2.1nm) and Fe–MSTF-3 (2.2 nm) is indeed too small to result in significant difference in the NLO signals. From the results, two conclusions can be drawn. First, at the iron oxide loading level lower than 1.5 at.%, no significant NLO response at 1064nm can be detected for sample without magnetic field treatment; Second, significant NLO response at 1064nm can be obtained at iron oxide contents as low as 0.83–1.45 at.% after heat treatment under 6 T magnetic field, but not at lower than around 0.8 at.%. That means, the off-resonance NLO property of the iron oxide incorporated MSTFs can be well-induced by the magnetic field heat treatment at certain Fe contents. Compared with the reported pure iron oxide thin films [5–7] which shown relative high χ(3) values from 10−11 to 10−8 esu at the laser wavelengths around 500 nm in the apparent absorption region, in this study, the sample Fe–MSTF-2 and Fe–MSTF-3 with greatly lower Fe loading amount (<1.5 at.%) after 6 T magnetic field heat treatment presented high χ(3) value at 10−10 esu magnitude order at 1064 nm which is far away from the absorption region of the incorporated iron oxide NPs ( Fig. 5 ).
To the best of our knowledge, this is the first report for the induced NLO property of the transition metal oxides NPs incorporated MSTFs by a high intensity magnetic field during heat treatment. Considering the superparamagnetic property of iron oxide NPs [22,23] which can be oriented by the magnetic field [24–26], we propose that the NLO performance generations were due to the orientation of the magnetic domains of the iron oxide NPs within MSTFs under the 6 T magnetic field during the thermal treatment (i.e. the magnetic orientation effect for the incorporated iron oxide NPs). The uniform orientation of the magnetic domains of the iron oxide NPs along the magnetic field direction [24–26] as induced by the magnetic field, is proposed to result in the significant nonlinear polarization of electric-magnetic field within the composite thin films under the incident laser irradiation and then the NLO performance generations. The more detailed and clear experimental evidences of the third-order NLO property generation of the materials by magnetic field heat treatment, though very difficult to obtain, are still under investigation.
In summary, we evidenced the generation of off-resonance third-order optical nonlinearities (χ(3) ~10−10 esu) of the iron oxide incorporated mesoporous silica thin films at 1064nm by the thermal treatment at 800 °C for 1 h under 6 T magnetic field, compared with the non-detecting of the NLO responses of the identical samples heat-treated without magnetic field by Z-scan technique. The third-order optical nonlinearity generation of the materials can be attributed to the oriented arrangement of the magnetic domains of the iron oxide NPs (i.e. the magnetic orientation effect for the incorporated iron oxide NPs), under the induction of a 6 T magnetic field along the magnetic field direction, which resulted in the significant nonlinear polarization of the composite thin films under the incident laser irradiation.
The works were supported by National Natural Science Foundation of China (Grant No. 20703055).
References and links
2. M. J. Bloemer, J. W. Haus, and P. R. Ashley, “Degenerate four-wave mixing in colloidal gold as a function of particle-size,” J. Opt. Soc. Am. B 7(5), 790 (1990). [CrossRef]
3. K. Fukumi, A. Chayahara, K. Kadono, T. Sakaguchi, Y. Horino, M. Miya, K. Fujii, J. Hayakawa, and M. Satou, “Gold nanoparticles ion-implanted in glass with enhanced nonlinear-optical properties,” J. Appl. Phys. 75(6), 3075 (1994). [CrossRef]
4. J. Gu, J. Shi, G. You, L. Xiong, S. Qian, Z. Hua, and H. Chen, “Incorporation of highly dispersed gold nanoparticles into the pore channels of mesoporous silica thin films and their ultrafast nonlinear optical response,” Adv. Mater. 17(5), 557–560 (2005). [CrossRef]
5. M. Ando, K. Kadono, M. Haluta, T. Sakaguchi, and M. Miya, “Large third-order optical nonlinearities in transition-metal oxides,” Nature 374(6523), 625–627 (1995). [CrossRef]
6. T. Hashimoto, T. Yamada, and T. Yoko, “Third-order nonlinear optical properties of sol–gel derived α- Fe2O3, γ- Fe2O3, and Fe3O4 thin films,” J. Appl. Phys. 80(6), 3184 (1996). [CrossRef]
7. H. Zhou, A. Mito, D. Kundu, and I. Honma, “Nonlinear optical susceptibility of Fe2O3 thin film synthesized by a modified sol-gel method,” J. Sol-Gel Sci. Technol. 19(1/3), 539–541 (2000). [CrossRef]
8. B. Yu, C. Zhu, and F. Gan, “Large nonlinear optical properties of Fe2O3 nanoparticles,” Physica E 8(4), 360–364 (2000). [CrossRef]
9. B. Yu, C. Zhu, F. Gan, X. Wu, G. Zhang, G. Tang, and W. Chen, “Optical nonlinearities of Fe2O3 nanoparticles investigated by Z-scan technique,” Opt. Mater. 8(4), 249–254 (1997). [CrossRef]
10. J. Gu, J. Shi, L. Xiong, H. Chen, and M. Ruan, “A new strategy to incorporate highly dispersed nanoparticles into the pore channels of mesoporous silica thin films,” Microporous Mesoporous Mater. 74(1-3), 199–204 (2004). [CrossRef]
11. P. Derango, M. Lees, P. Lejay, A. Sulplice, R. Tounier, M. Ingold, P. Germi, and M. Pernet, “Texturing of magnetic materials at high temperature by solidification in a magnetic field,” Nature 349(6312), 770–772 (1991). [CrossRef]
12. S. Uji, H. Shinagawa, T. Terashima, T. Yakabe, Y. Terai, M. Tokumoto, A. Kobayashi, H. Tanaka, and H. Kobayashi, “Magnetic-field-induced superconductivity in a two-dimensional organic conductor,” Nature 410(6831), 908–910 (2001). [CrossRef] [PubMed]
14. H. Wang, X. Ma, Y. He, S. Mitani, and M. Motokawa, “Enhancement in ordering of FePt films by magnetic field annealing,” Appl. Phys. Lett. 85(12), 2304 (2004). [CrossRef]
15. D. Zhao, P. Yang, N. Melosh, J. Feng, B. Chmelka, and G. Stucky, “Continuous mesoporous silica films with highly ordered large pore structures,” Adv. Mater. 10(16), 1380–1385 (1998). [CrossRef]
16. F. Cui, Z. Hua, X. Cui, L. Guo, C. Wei, W. Bu, and J. Shi, “Au nanoparticles incorporated mesoporous silica thin films with a high Au content: preparation and third-order optical non-linearity,” Dalton Trans. (15): 2679–2682 (2009). [CrossRef] [PubMed]
17. D. G. Barton, M. Shtein, R. D. Wilson, S. L. Soled, and E. Iglesia, “Structure and electronic properties of solid acids based on tungsten oxide nanostructures,” J. Phys. Chem. B 103(4), 630–640 (1999). [CrossRef]
18. R. A. Ganeev, A. I. Ryasnyanskiy, and T. Usmanov, “Optical and nonlinear optical characteristics of the Ge and GaAs nanoparticle suspensions prepared by laser ablation,” Opt. Commun. 272(1), 242–246 (2007). [CrossRef]
19. M. Sheik-Bahae, A. A. Said, T. H. Wei, D. J. Hagan, and E. W. Van Steyland, “Sensitive measurement of optical nonlinearities using a single beam,” IEEE J. Quantum Electron. 26(4), 760–769 (1990). [CrossRef]
20. R. Ganeev, A. Ryasnyansky, A. Stepanov, and T. Usmanov, “Nonlinear optical susceptibilities of copper- and silver-doped silicate glass in the ultraviolet range,” Phys. Status Solidi 238(2), R5–R7 (2003) (b). [CrossRef]
21. H. S. Kim, M. H. Lee, N. C. Jeong, S. M. Lee, B. K. Rhee, and K. B. Yoon, “Very high third-order nonlinear optical activities of intrazeolite PbS quantum dots,” J. Am. Chem. Soc. 128(47), 15070–15071 (2006). [CrossRef] [PubMed]
22. S. Trudel, C. Jones, and R. Hill, “Magnetic properties of nanocrystalline iron oxide/amorphous manganese oxide nanocomposite thin films prepared via photochemical metal-organic deposition,” J. Mater. Chem. 17(21), 2206 (2007). [CrossRef]
23. A. Fuertes, M. Sevilla, S. Alvarez, T. Valdes-Solis, and P. Tartaj, “Templated synthesis of mesoporous superparamagnetic polymers,” Adv. Funct. Mater. 17(14), 2321–2327 (2007). [CrossRef]
24. Y. Sahoo, M. Cheon, S. Wang, H. Luo, E. Furlani, and P. Prasad, “Field-directed self-assembly of magnetic nanoparticles,” J. Phys. Chem. B 108(11), 3380–3383 (2004). [CrossRef]
25. J. Lin, W. Zhou, A. Kumbhar, J. Fang, E. Carpenter, and C. O’Connor, “Gold-coated iron (Fe@Au) nanoparticles: Synthesis, characterization, and magnetic field-induced self-assembly,” J. Solid State Chem. 159(1), 26–31 (2001). [CrossRef]
26. M. A. Correa-Duarte, M. Grzelczak, V. Salgueiriño-Maceira, M. Giersig, L. M. Liz-Marzan, M. Farle, K. Sierazdki, and R. Diaz, “Alignment of carbon nanotubes under low magnetic fields through attachment of magnetic nanoparticles,” J. Phys. Chem. B 109(41), 19060–19063 (2005). [CrossRef]