1. Introduction

Periodic structures of photonic crystals allow the control of the propagation of light. Therefore, the photonic crystals have attracted much attention as a new optical material for manipulating light such as an optical fiber. Thomas et al. has progressively demonstrated one-, two- and three-dimensional photonic crystal structures based on self-assembly of block copolymers [1–4]. They also demonstrated the application of block copolymer photonic crystals to resonators in order to provide spectrally selective feedback for lasing [5]. Our group has been developing photonic crystals made of ultra-high-molecular-weight block copolymers, and succeeded in fabrication of giant-grain photonic crystals with highly ordered microdomain structures from the dilute solutions of the copolymers and fixation of these structures. In addition, we demonstrated laser emission achieved by a combination of laser dye and the resonator fabricated with the giant grains in the solutions. These results will be reported elsewhere.

On the other hand, gold nanoparticles have been attracting attention as optical nonlinear materials [6]. These properties are expected to be used for fabrication of optical switching devices and frequency converters [7]. Realization of optical switching functions and high efficiency need a large third-order optical susceptibility. Many works reported that the gold nanoparticles dispersed in glass or film possessed large third-order optical susceptibilities [8,9]. More recently, the effect of gold nanoparticle arrays for the enhancement of third-order optical susceptibilities was indicated by Shen et al. [10], Ning et al. [11] and Wang et al. [12]. They fabricated the periodic gold nanoparticle arrays by using nanosphere lithography with a self-assembly pattern of polystyrene nanospheres prepared by drop coating as a mask, and observed a large nonlinear optical response. Other advantage of gold nanoparticles is its simplicity to coat them by various kinds of polymer, and many researches about control of nanoparticle location in block copolymer template were reported [13,14]. Especially, Bockstaller et al. selectively sequestered gold nanoparticles into microdomain structures and observed the change in absorption spectra by varying the particle density in the microdomains [15].

The aim of this research is to develop a nonlinear optical material having a photonic crystal structure for the enhancement of optical nonlinearities using the periodic microdomain structure of block copolymers. Our block copolymer-based photonic crystals themselves do not exhibit optical nonlinearities. Therefore, nanometer-sized metal particles must be intentionally introduced into the photonic crystal structures to attain optical nonlinearity. Our final goal is to construct new optical switching devices that consist of gold nanoparticles dispersed into the selective domains of block copolymer periodic structures (photonic crystals). We have already succeeded in demonstrating switching function, and will report it in near future. The switching function is generated by the enhancement of their nonlinear optical properties reported in this paper. The nonlinear optical properties can be enhanced by two effects: First, the periodic gold nanoparticle arrays localize the nanoparticles at the antinodes of the standing resonant wave. Second, the interparticle coupling between the gold nanoparticles confined in the selective microdomains induces the electric charge concentration [16,17] and the optical nonlinearities can be enhanced (local field effect). In addition, we aim to enhance the optical nonlinearities by adjusting the domain spacing of the template polymers with respect to the surface plasmon resonance (SPR) peak, which enhances the resonance behavior at the plasmon frequency.

2. Experimental

By sequential living anionic polymerization we synthesized a poly(styrene-b-methyl methacrylate) (PS-b-PMMA) diblock copolymer in tetrahydrofuran as solvent at -78 °C and used as the template for the site-selective introduction of gold nanoparticles. Polymerization of styrene was initiated by sec-butyllithium and allowed to proceed for 2 h. Subsequently, methyl methacrylate was added to form PS-b-PMMA diblock copolymer and allowed to proceed for 24 h, followed by the termination with acidic methanol. The number-average molecular weight M _n, the polydispersity index (PDI) and the volume fraction of PS (ϕ_PS) of PS-b-PMMA were 3.2×10⁵ g/mol, 1.61 and 31 vol.%, respectively. Similarly, poly(styrene-b-tert-butyl methacrylate) (PS-b-PtBMA) (M _n=6.8×10⁵ g/mol, PDI=1.12, ϕ_PS=46 vol.%) used as a template polymer, polystyrene homopolymer (PS) (M _n=7.9×10⁵ g/mol, PDI=1.78) used as a matrix for uniform dispersion of gold nanoparticles, and thiol-terminated polystyrene (PS-SH) (M _n=6.8×10³ g/mol, PDI=1.09) employed as a coating polymer of gold nanoparticles were synthesized. Finally we synthesized poly(methyl methacrylate) homopolymer (PMMA) used for blends by radical polymerization with toluene as solvent and 2,2′-azobis(isobutyronitrile) as initiator. Polymerization of methyl methacrylate was allowed to proceed at 60 °C under nitrogen atmosphere for 2 h. The M _n and PDI of PMMA was 1.9×10⁴ g/mol and 1.3, respectively.

The synthesis of polystyrene-coated gold nanoparticles (PS-Au) was accomplished using a two-phase system consisting of toluene and water [18]. The average size of PS-Au was evaluated to be 3.5 nm by transmission electron microscopy (TEM) using a Hitachi HF-2000 FE-TEM. We observed the SPR peak of PS-Au, implying the resonant behavior at the plasmon frequency of free electrons confined in the Au nanoparticles, in the visible absorption spectra obtained with a SHIMADZU MultiSpec-1500 spectrophotometer. It was confirmed that the SPR of PS-Au was peaked at 527 nm in toluene as shown later in Fig. 2.

To prepare the nanocomposite films (PS or block copolymer films with Au nanoparticles), 2 wt.% PS or one of the block copolymers was first dissolved in benzene containing PS-Au. The weight ratio of PS-Au to the polymer weight was fixed to 0.5 wt.%. In this way, we prepared four different solutions: PS/PS-Au, PS-b-PMMA/PS-Au, and PS-b-PMMA/PMMA/PS-Au (two PS-b-PMMA/PMMA blends with ϕ_PS=3 and 10 vol.%) solutions. Nanocomposite films were obtained by casting the solutions at 25 °C by slow evaporation of the solvent for a month and subsequently dried under vacuum for 6 h to remove any residual solvent. Thus, we obtained 100 μm-thick cast films of various polymer/PS-Au nanocomposites. We also prepared an as-cast film without PS-Au as a control sample from a 2 wt.% PS-b-PMMA/benzene solution.

For the TEM observation, the as-cast film samples were subjected to ultramicrotoming into ultrathin sections of 90 nm thickness with glass knives by using a Leica EM UC6 ultramicrotome. The PS domains in the thin sections were selectively stained with ruthenium tetroxide (RuO₄). The SPR peak wavelength of each sample was determined by the visible spectrophotometry also with a SHIMADZU MultiSpec-1500 spectra. The optical nonlinearities were measured using the Z-scan technique [19], which enable us to obtain third-order optical susceptibility (χ⁽³⁾) for the four samples. The Z-scan measurements were performed using a Ti : Sapphire mode lock laser operated at the wavelength of 800 nm with the pulse width of 20 fs and frequency of 62.5 MHz. This laser has a broad spectrum with the intensity maximum at 800 nm, thus the intensity at the wavelength of 527 nm, where the SPR peak of PS-Au exists, is strong enough to induce the SPR. The measurement was calibrated using ZnSe as a standard [20].

Separately, we also prepared PS-b-PtBMA/PS-Au toluene solutions by the same procedure described above. In these solutions, polymer concentration was varied from 7 to 26 wt.% and the concentration of PS-Au was fixed to 0.1 wt.%. The PS-b-PtBMA/PS-Au solution was sandwiched between two glass plates for the following measurements. The peak wavelengths of the iridescent color, attributed to the Bragg reflection from the periodic microphase-separated structures in these solution samples, were determined by the visible spectrophotometry. The domain spacings were calculated from the reflection peak wavelengths using the Bragg equation and the volume-averaged refractive index using the values for toluene (n=1.50), PS (n=1.59) and PtBMA (n=1.46). The optical nonlinearities of the solution samples were also estimated using the Z-scan technique.

3. Results and discussion

Not only the microphase-separated structures of the polymer/PS-Au nanocomposite as-cast films but also the distribution of the gold nanoparticles were investigated by TEM, as shown in Fig. 1. The PS/PS-Au film (hereafter denoted as ϕ_PS100) had a homogeneous and random distribution of gold nanoparticles, as seen in Fig. 1(a). Since the PS domains of the block copolymer samples were selectively stained by RuO₄, the darker domains in their images correspond to the PS domains. The TEM images in Fig. 1(b)–(d) show the dependence of microdomain morphology on the PS volume fraction in the film samples. The PS-b-PMMA/PS-Au film (ϕ_PS31, Fig. 1(b)) shows a lamellar microdomain structure with a high regularity. The PS-b-PMMA/PMMA/PS-Au film (ϕ_PS10, Fig. 1(c)) with the volume fraction of PS at 10 vol.% shows a spherical microdomain structure with a cubic arrangement of the spheres. The PS-b-PMMA/PMMA/PS-Au film (ϕ_PS3, Fig. 1(d)) with the volume fraction of PS at 3 vol.% shows also a spherical but sparse microdomain structure without regular arrangement of spheres. Careful TEM observations of the TEM images in Fig. 1 allowed us to conclude that the PS-coated Au nanoparticles (PS-Au) were located only in the PS domains in every sample.

We measured by visible spectrophotometry absorption spectra of the four samples used for the TEM observations together with two control samples: a PS-b-PMMA film without PS-Au and a toluene solution of PS-Au. Figure 2 clearly shows that the PS-b-PMMA film without PS-Au (spectrum (f)) has no absorption peak while the four polymer/PS-Au nanocomposite films (spectra (a)–(d)) have the absorption peaks near the wavelength (527 nm) of the SPR peak of a toluene solution of PS-Au (spectrum (e)). Thus these four absorption peaks can be attributed to the SPR peaks of the Au nanoparticles in the films. While the SPR peak wavelength of 527 nm observed for ϕ_PS100 and φ_PS31 (Fig. 2(a) and (b), respectively) coincides with the wavelength at the SPR peak of PS-Au, those of ϕ_PS10 (Fig. 2(c)) and ϕ_PS3 (Fig. 2(d)) were shifted to 537 and 542 nm, respectively. These red shifts indicate the local field effect, which was caused by the interparticle coupling due to the short interparticle distances between the Au nanoparticles confined in the small spherical microdomain space [15,21]. It is worth noting that the PS-Au nanoparticles can be uniformly distributed in the PS lamellar domains because the SPR peak of ϕ_PS31 coincides with the wavelength at the SPR peak of PS-Au as explained by Bockstaller et al. [15]. In consideration of the structural observation in Fig. 1 and the red shifts of the SPR peaks in Fig. 2, four different types of sample films containing in terms of microdomain structures and spatial arrangements of Au nanoparticles have been prepared: Type (a) is PS/PS-Au (ϕ_PS100) where Au nanoparticles are homogeneously dispersed. Type (b) is PS-b-PMMA/PS-Au (ϕ_PS31) where Au nanoparticles are preferentially located in the lamellar PS microdomains of relatively large volume fraction (31 %) giving rise to the periodic arrangement of the Au nanoparticles. It is expected from this type of nanocomposite that the periodic arrays of the Au nanoparticles in the photonic crystal structure would provide positive effects [10–12] on their optical nonlinearities. Type (c) is PS-b-PMMA/PMMA/PS-Au (ϕ_PS10) where Au nanoparticles are confined and concentrated in the spherical PS microdomains with a small volume fraction (10 %) giving rise to the local field effect by interparticle coupling between the neighboring Au nanoparticles in addition to the effect of the periodic arrays of the Au nanoparticles controlled by the regular arrangement of the spherical PS microdomains. Type (d) is PS-b-PMMA/PMMA/PS-Au (ϕ_PS3) where Au nanoparticles are confined in the irregularly arranged spherical PS microdomains with a very small volume fraction (3 %) resulting in the enhanced local field effect. However, no effect of periodic arrangement of the Au particles can be expected.

 

Fig. 1. TEM images of polymer/PS-Au nanocomposite films: (a) PS/PS-Au (ϕ_PS100), (b) PS-b-PMMA/PS-Au (ϕ_PS31), (c) PS-b-PMMA/PMMA/PS-Au (ϕ_PS10), (d) PS-b-PMMA/PMMA/PS-Au (ϕ_PS3). Scale bar is 100 nm.

Download Full Size | PDF

 

Fig. 2. Visible absorption spectra of the polymer/PS-Au nanocomposite films: (a) PS/PS-Au (ϕ_PS100), (b) PS-b-PMMA/PS-Au (ϕ_PS31), (c) PS-b-PMMA/PMMA/PS-Au (ϕ_PS10) and (d) PS-b-PMMA/PMMA/PS-Au (ϕ_PS3). Control samples are also shown: (e) PS-Au in toluene and (f) PS-b-PMMA without PS-Au. The arrows indicate the SPR peak positions: (a) 527 nm, (b) 527 nm, (c) 537 nm, (d) 542 nm and (e) 527 nm. A SPR peak shift is observed in (c) and (d).

Download Full Size | PDF

Typical peak-to-valley profiles, indicating a negative value of the nonlinear refractive index n ₂, were obtained for all of the four polymer/PS-Au nanocomposite films by the Z-scan measurement, which determines the value of χ⁽³⁾. However, the control film without PS-Au, which has no absorption peak as mentioned above, did not exhibit the peak-to-valley one. In other words, the film without PS-Au, i.e., the matrix media to distribute PS-Au has no optical nonlinearity. Hence the nonlinearity we observe here is attributed to PS-Au. This holds true for the solution systems that we will describe later. It should be noted that despite that this measurement was performed using a laser with the intensity maximum at 800 nm, we could obtain χ⁽³⁾ values for the Au nanoparticles whose SPR peak was observed at 527 nm. This result indicates that the intensity of this laser at the wavelength of 527 nm is strong enough to induce the SPR. Figure 3 shows the values of χ⁽³⁾ determined by the Z-scan measurement of the four film samples mentioned above. The data indicated by the horizontal lines on the error bars represent the value of χ⁽³⁾ that were independently measured by the Z-scan technique for three times and the filled circles denote their average values. The average values of χ⁽³⁾ obtained for ϕ_PS100, ϕ_PS31, ϕ_PS10 and ϕ_PS3 are 4.15×10^-11, 7.68×10^-11, 8.15×10^-11 and 5.82×10^-11 esu, respectively. The result that χ⁽³⁾ of ϕ_PS3 (Type (d)) is greater than that of ϕ_PS100 (Type (a)) implies the enhanced optical nonlinearity by the local field effect caused by the electric charge concentration due to the confinement of the Au nanoparticles within the spherical microdomains with a small volume fraction [16,17]. Next, we compare ϕ_PS10 (Type (c)) and ϕ_PS3 (Type (d)). The former has the periodic arrangement of the Au nanoparticles but less concentration (local field effect) than the latter. Therefore, the higher value of χ⁽³⁾ for ϕ_PS10 than for ϕ_PS3 must have been resulted from the enhancement of optical nonlinearities by the effects of periodic Au nanoparticle arrays. Comparison of the difference of the χ⁽³⁾ values between ϕ_PS31 (Type (b)) and ϕ_PS100 (Type (a)) with that between ϕ_PS3 (Type (d)) and ϕ_PS100 enables us to compare the effects of the periodic arrangement of Au nanoparticles and the interparticle coupling on the enhancement of the optical nonlinearities because all the samples contain 0.5 wt.% PS-Au. Obviously, the periodic arrays of Au nanoparticles exhibited grater effect than the interparticle coupling. The sample such as ϕ_PS10 (Type (c)) possessing both periodic arrays and coupling shows the greatest enhancement of optical nonlinearities. Note that these results were obtained and could be valid for the systems containing 0.5 wt.% PS-Au. The local field effect could be more dominant than the effect of the periodic arrangement for the system contains Au nanoparticles with much higher concentration.

 

Fig. 3. The values of the third-order susceptibility, χ⁽³⁾, of ϕ_PS100, ϕ_PS3, ϕ_PS31 and ϕ_PS10. Three horizontal lines on each error bar represent the values of χ⁽³⁾ obtained for three independent Z-scan measurements. The filled circles denote the average values of χ⁽³⁾. The highest value of χ⁽³⁾ was observed for ϕ_PS10, while the lowest for ϕ_PS100.

Download Full Size | PDF

On the basis of the above-mentioned results, we next conducted the study on the effect of the domain spacing of the template polymer for the optical nonlinearities. Figure 4(a) shows the reflection peak positions of the PS-b-PtBMA/PS-Au solutions sandwiched by two glass plates in which polymer weight fraction was varied from 10 to 26 wt.% were measured by visible spectrophotometry. The error bars represent the full width at half maximum of the reflection peak. We guessed the microdomain morphology of the solutions as a lamella structure because the volume fraction of PS in PS-b-PtBMA is 0.46. The solutions of the polymer weight fractions from 7 to 9 wt.% did not exhibit any reflection peak. Only the solutions with the polymer concentration more than 10 wt.% exhibited a reflection peak and seemed to have a highly-ordered microphase-separated structure. The position of the reflection peaks shifts toward a longer wavelength with increasing the polymer weight fraction because of the increase in the domain spacing induced by an increase in the segregation power with ϕ_PχN [22], where ϕ_P is polymer volume fraction, N is degree of polymerization and χ is Flory-Huggins interaction parameter. Moreover, it is well-known that χ depends on temperature. The value of χ increases with decreasing temperature for such a system exhibiting a upper critical solution temperature (UCST) type phase behavior as PS-b-PMMA or PS-b-PtBMA, in other words, ϕ_PχN increases with decreasing temperature [22]. We measured the temperature dependence of the reflection peak wavelength for the PS-b-PtBMA/PS-Au samples from 22 to 26 wt.% by decreasing temperature. (We obtained more results on the temperature dependence for the Z-scan measurement around the wavelength of 527 nm, the wavelength of the SPR peak of PS-Au, as shown below in Fig. 4(c).) Indeed, we observed the red shift of the reflection peaks with decreasing temperature for these samples and added the results in Fig. 4(a) as shown by triangles and squares. The reflection peak wavelength from the 24 wt.% solution was successfully observed at 527 nm. In addition, Fig. 4(a) exhibits a linear increase in the reflection peak wavelength with the polymer weight fraction, so the lamellar morphology was considered to be conserved for all the polymer weight fraction in this experiment [22]. Figure 4(b) shows the polymer concentration dependence of χ⁽³⁾ obtained by the Z-scan measurement (In Fig. 4(b), three horizontal lines on the error bars denote χ⁽³⁾ values which were obtained by three independent Z-scan measurements and the filled circles and the triangles denote the average values of χ⁽³⁾). The 10 wt.% solution was inhomogeneous exhibiting some parts with an iridescent color and the others without iridescent color. These two selected areas with and without iridescent color were examined for optical nonlinearities by the Z-scan technique. The value of χ⁽³⁾ of the 22, 24 and 26 wt.% solutions in Fig. 4(b) were the results obtained at room temperature for the samples that have the reflection peak wavelength at 513, 524 and 543 nm, respectively (shown by the filled circles in Fig. 4(a)). The result of 10 wt.% solution in Fig. 4(b) shows that χ⁽³⁾ measured at the part with the iridescent color (the filled circle at 10 wt.% in Fig. 4(b)) is larger than that at the part without iridescent color (the triangle at 10 wt.% in Fig. 4(b)). Similarly, the solutions with more than 10 wt.% polymer concentration with iridescent color all exhibit a lager χ⁽³⁾ than those with less than 10 wt.% polymer concentration without iridescent color. This tendency indicates that an increase in regularity, i.e., both an increase in number of stacking layers and a decrease in lattice distortion, of the microphase-separated structure of the template polymer and the gold nanoparticles arrays have great influences to enhance the optical nonlinearities of the Au nanoparticles. The values of χ⁽³⁾ increased with increasing polymer concentration as shown in Fig. 4(b), while we observed the decrease of χ⁽³⁾ at 26 wt.%. We next discuss this phenomenon. In Fig. 4(c), we replotted χ⁽³⁾ as a function of the apparent domain spacing because we want to investigate the effect of the domain spacing on χ⁽³⁾. The apparent domain spacing that is defined as 2nd, where n is the refractive index and d is the domain spacing of the microdomain structures in the template polymer solution, is apparently the same value as the reflection wavelength. We observed that χ⁽³⁾ increased with an increase in the apparent domain spacing up to 527 nm, and then decreased above 527 nm. The equivalent tendency can be seen at 26 wt.% polymer concentration in Fig. 4(b). The apparent domain spacing with the maximum value of χ⁽³⁾ was observed at 527 nm, that is noticeably the same wavelength as the SPR peak of PS-Au. Therefore, we compare the values of χ⁽³⁾ with the SPR spectrum by normalizing the SPR intensity after the subtraction of background coming from band to band transition and superposing the normalized SPR spectrum onto χ⁽³⁾ to satisfy the condition that the peak position and the intensity in the tail region of the normalized SPR spectrum match those of χ⁽³⁾ in Fig. 4(c). Here, the domain spacing of the template polymer was calculated from the reflection wavelength using the Bragg equation, and is shown as the upper axis in Fig. 4(c). Here we used as n the volume averaged refractive indices of toluene solvent (n=1.50), PS (n=1.59) and PtBMA (n=1.46). We found that the dependence of χ⁽³⁾ on the apparent domain spacing is in good agreement with the SPR peak profile of PS-Au. Surprisingly, the values of χ⁽³⁾ at 524, 527 and 531 nm are much higher than the SPR peak intensity, which will hereinafter be discussed in detail. We consider that this result, except the wavelength region from 524 to 531 nm, is understood as follows. In a previous work, Uchida et al. reported that the value of χ⁽³⁾ increased with varying the pumping wavelength [23]. They reported using the pumping light with the wavelength where the higher SPR intensity was observed, i.e., the light with the wavelength closer to the SPR frequency, the optical nonlinearities were more enhanced as Hache et al. theoretically explained [24]. It should be reminded that in the strategy of this study a sophisticated effect of the template photonic crystal that the propagating light in the media is trapped by its periodic structure was intentionally appended for the enhancement of the optical nonlinearities of PS-Au doped in such a structure. The group velocity extremely decays at the photonic band edge. The optical nonlinearities are enhanced by the slower group velocity at the photonic band edge [25]. Our sample possesses a photonic band gap, i.e., the light with the same wavelength of its photonic band edge travels within the sample for a longer time. It is hence expected that the optical nonlinearities increases as a photonic band gap edge of the sample approaches closer to the SPR wavelength because the trapped light by photonic crystals behaves just like the pumping light reported by Uchida et al [23]. This effect is revealed by the fact that the extremely high values of χ⁽³⁾ are very noticeable at 524 nm (d=174 nm), 527 nm (d=175 nm) and 531 nm (d=176 nm). Yang et al. [26] found the interparticle linkage effect of Au nanoparticles, i.e., χ⁽³⁾ of “Au nanochains” composed of four or five particles becomes three times higher than that of isolated Au nanoparticles. Shen et al. [10] reported that χ⁽³⁾ of the Au nanoparticles in the well-ordered 2D arrays is about six times larger than that in randomly distributed spheroidal clusters. In contrast to the 2D array of nanoparticles in the thin film, the Au nanoparticles in our system have a random distribution of Au nanoparticles within a single lamellar microdomain while the microdomains align periodically in the direction perpendicular to the lamellae. Then the value of χ⁽³⁾ became about eight times larger by matching this periodicity to the SPR peak. This effect was not found by Uchida et al either. We consider that the enhancement by trapping light within the template polymer arises increasingly around the SPR peak wavelength. It should definitely be mentioned that the fabrication of the Au nanocomposite with the photonic band gap closer to the SPR peak wavelength gives rise to further gaining of the χ⁽³⁾ optical nonlinearities. Their application as optical switching devices by the prominent enhancement of optical nonlinearities, which was experimentally evidenced in this paper, is now in progress, and will be reported in near future.

 

Fig. 4. Polymer concentration dependence of the reflection peak wavelength (a) and χ⁽³⁾ (b) of the PS-b-PtBMA/PS-Au solutions, and the comparison of χ⁽³⁾ as a function of the apparent domain spacing of the template polymer solutions with the normalized SPR spectrum of PS-Au (c). (a) Error bars represent the full width at half maximum of the reflection peaks. The marks (triangles and squares) indicate the peak wavelengths obtained by temperature variations at the corresponding polymer concentration from 22 to 26 wt.%. (b) The horizontal lines on the error bars represent the values of χ⁽³⁾ that were determined by three Z-scan measurements and the filled circles and the triangles denote the average values of χ⁽³⁾. The upper point (filled circle) at 10 wt.% indicates the χ⁽³⁾ measured at the part with the iridescent color and the lower point (triangle) indicates that of the part without iridescent color. The results of 22, 24 and 26 wt.% were the results obtained at room temperature for the samples that have the reflection peak wavelength at 513, 524 and 543 nm, respectively. (c) The SPR spectrum intensity was normalized after the subtraction of background coming from band to band transition and surperimposed onto the values of χ⁽³⁾. The domain spacing was calculated from the reflection wavelength using the volume averaged refractive index and is shown as the upper axis. Three horizontal lines on each error bars represent the χ⁽³⁾ of the three independent Z-scan measurements and the average values of χ⁽³⁾ is denoted by the filled circles.

Download Full Size | PDF

4. Conclusion

In summery, we investigated the enhancement of the optical nonlinearities of gold nanoparticles by using the periodic microphase-separated structure of a diblock polymer photonic crystal as a template for dispersion of these nanoparticles. Selective introduction of gold nanoparticles into microphase-separated structures caused the local field effect and the effect of periodic gold nanoparticle arrays. These effects enhanced the optical nonlinearities. In addition, we investigated the dependence of the domain spacing of template polymer on χ⁽³⁾ by using the PS-b-PtBMA/PS-Au solutions. It was indicated that as the reflective peak wavelength of the iridescent color correlated with the domain spacing approach the SPR peak wavelength, χ⁽³⁾ exhibited a higher value. Crucially, the extreme enhancement of the optical nonlinearities was observed when the template polymer possessed a photonic band edge at nearly SPR peak wavelength of gold nanoparticles in PS domains.