1. Introduction

The spatial self-phase modulation (SSPM) as a typical process of light intensity and phase modulation has been systematically investigated decades ago [1]. Due to the spatial refraction index gradient excited by laser beam in the photosensitive material, the transmitted beam evolves into multiple concentric diffraction rings. The distribution of the intensity dependent refractive index could also affect other light passing through the irradiated region. According to this property, a light-control-light system for light transmission modulation has been proposed recently. A two-color all-optical switching with superb performance is proposed by Wu et al. firstly [2]. Such all-optical switching realizes modulating a weak beam by a strong beam and has attracted the attention of many researchers [2–10]. Its mechanism has been explained as cross-phase modulation [3] or spatial cross-phase modulation (SXPM) [5,8,11]. The controlling light is used to excite and modulate the diffraction rings of the controlled light propagating through the same region of the nonlinear optical material by adjusting its intensity. The input information induced by the controlling light can be transmitted to the signal light, which could be applied in all-optical information conversion [4,9,11].

On the other hand, the effect of the light modulation and information conversion would be affected by various factors. For instance, the diffraction pattern of controlled light with different wavelength passing through the same region exhibit different size and ring number. The cross angle of the two beams affects the shape of the diffraction pattern of the controlled light, which would be distorted compared with that of controlling light in previous reports [4,5,8]. Most of all, those reported SXPM were realized in the liquid samples [2–9,11] and the heat of the laser beam would cause the thermal convection of the liquid [12–14]. The diffraction pattern of the beam passing through the liquid sample become squeezed from the top instead of remaining perfectly round as a consequence of the nonaxis-symmetrical thermal convection of the liquid [12,15,16]. The collapse effect of the concentric diffraction rings leads to the distortion of the circular pattern and the convection of the liquid sample makes the pattern unstable [2,4–9,11,12]. For better all-optical information conversion effect, the thermal convection should be eliminated as much as possible. Preparing the sample in solid state provides a method to eliminate the thermal convection completely. However, the heat of the laser beam might cause the deformation of sample surface, resulting in the distortion of the diffraction rings. Liquid crystal is a transition state between liquid and crystalline with both liquid mobility and crystal anisotropy. SSPM as a simple and effective method is widely used in investigating the optical nonlinearity of the liquid crystals. The light induced thermal convection of the liquid crystals sample could be suppressed, and the diffraction rings obtained by the SSPM method are circular symmetric [17,18] and more stable. The previous study of our group about the SSPM in azo dye doped nematic liquid crystals also observed the perfectly round diffraction rings [19]. The azo dye doped liquid crystals with various advantages of low cost, easy preparation and long-term preservation could be a choice to realize the SXPM without collapse effect of the diffraction patterns, providing potential applications prospect in designing advanced optical devices for all-optical switching, modulation and information conversion.

In this paper, we investigated the modulation effect of a controlled light in the azo dye doped liquid crystals by means of the SXPM method. The all-optical switching was demonstrated only when the controlling light (532 nm) and the controlled light (633 nm) were merged together. The diffraction pattern of the controlled light exhibited good circular symmetry. When there was an intersection angle between the two lights, the controlled light also diffracted into multiple circular fringes but the pattern differed from the diffraction pattern of the controlling light significantly.

2. Experimental

The sample was the nematic liquid crystals of pentylcyanobiphenyl (5CB, Merck) doped with the azo-dye disperse red 1 (DR1, Aldrich) at a concentration of about 1 wt.%. The preparation process was listed as follows: the mixture of the liquid crystals and the azobenzene dye was magnetically stirred at room temperature in the dark until the DR1 was dissolved and then injected into an empty planar aligned liquid crystals cell with a cell gap of 50 μm. A CW Nd: YAG laser with a wavelength of 532 nm which was located at the absorption band was employed as the controlling light. The controlling light was focused on the center of the cell by a lens (f = 100 mm), and the intensity was modulated by an attenuation slice. The azo dye doped liquid crystals sample was placed vertically after the lens and the distance between the front surface of the cell and the lens was set at 105 mm. A He-Ne laser with a wavelength of 633 nm which was out of the absorption band was employed as the controlled light. The power of 633 nm light passing through the lens and the sample was 0.7725 times its initial power. The controlling and controlled lights were both linearly polarized in the vertical direction. Two beams were merged together and a cut-off filter was used to block the controlling light for better observation. The experimental setup was displayed in Fig. 1. The transmitted light patterns were observed on a white screen suspended behind the liquid crystals sample, and captured by a digital camera (Canon, EOS600D). The distance between the white screen and the cell was 110 cm.

 

Fig. 1. Schematic illustration of the experimental setup. Inset: the photograph of the azo dye doped liquid crystals and the rubbing direction of the liquid crystals cell.

Download Full Size | PDF

3. Results and discussion

The modulation effect of the initial 633 nm light controlled by 532 nm light was investigated and the results were shown in Fig. 2. The power of the 532 nm light was increased from 0 to 40 mW with the increments of 5 mW. The power of the 633 nm light was set at 10 mW and only a red spot could be observed without the irradiation of the 532 nm light, indicating that there was no SSPM effect of the controlled light with this power. As the 532 nm light with the power of 5 mW irradiated on the sample, the spot of the 633 nm light changed immediately. The shapes of the two transmitted lights were almost the same. With the power of the 532 nm light increasing to 20 mW, the transmitted light evolved into a spot surrounded by one ring. Two diffraction rings were excited when the power of the 532 nm light was set to 30 mW. The size of the diffraction pattern was also expanded with the enhancement of the light power. The light pattern of the 633 nm light evolved accompanied with the conversion of the pattern of the 532 nm light. The ring number and the size of the diffraction pattern of the 633 nm light increased with the power of the 532 nm light rising. However, the ring number of 633 nm was not always the same as that of 532 nm light. In this work, the 633 nm light diffracted into two rings with the maximum power of the 532 nm light (48 m W). The liquid crystals 5CB have three phased states (crystal, nematic and isotropic phase) under different environment temperature. The light-control-light system worked well in all the phase states.

 

Fig. 2. The photographs of the overlapped light patterns of the 633 nm light (10 mW) and the 532 nm light with the power of 0, 5, 10, 15, 20, 25, 30, 35, 40 and 48 mW. The distance between the liquid crystals sample and the white screen was approximately 110 cm. Scale bars, 10 mm.

Download Full Size | PDF

It is well known that the azo molecules have two kinds of isomers with different molecular shapes and polarity: rod-like tans form and bent cis form. [20] When the linearly polarized 532 nm light irradiated on the sample, the azo molecules underwent the isomerization and aligned in the perpendicular direction of the polarization of the controlling light, resulting a photoinduced birefringence. Diffraction rings of the controlling light could be obtained with its power over the excitation threshold of the sample. When a 633 nm light passed through the irradiated sample, the changed reflective index established by the controlling light would also cause a phase shift on the controlled light and the controlled light diffracted into concentric rings. The clear diffraction patterns of the 633 nm light obtained by the cut-off filter were shown in Fig. 3 (a). Unlike the diffraction pattern of the 633 nm light modulated in the liquid sample (see the Supplement 1), all the light patterns in Fig. 3 (a) remained perfectly round. The circular symmetry of the diffraction rings of the controlled light was preserved well, indicating that the thermal convection induced by the heat of the controlling laser beam was suppressed successfully in liquid crystals sample. The relationship between the size of the patterns of the two lights and the power of 532 nm light was analyzed in Fig. 3 (b). Apparently, a linear dependence between the diameter of the light pattern and power could be observed. After fitting, the slopes of the 532 nm light and 633 nm light were 0.9928 and 0.6515, respectively. The diameter of the circular diffraction pattern of the controlled light increased linearly with the enhancement of the controlling light power.

 

Fig. 3. (a) The photographs of the light patterns of the 633 nm light (10 mW) under the irradiation of the 532 nm light with the power of 0, 10, 20, 30, 40 and 48 mW obtained by the cut-off filter. (b) The variation of the diameters of the patterns of the two lights with the enhancement of the power of 532 nm light. The distance between the liquid crystals sample and the white screen was approximately 110 cm. Scale bars, 10 mm.

Download Full Size | PDF

The response speed of this light-control-light system would be an important parameter for the real applications. For further investigating the response speed of the all-optical switching, A position on the outermost ring of the diffraction patterns of the 633 nm light was chosen and a photodetector connected with an oscilloscope was placed there to record the light signal. When the 532 nm light was turned off, the 633 nm light did not diffract and no light signal could be detected. When the 532 nm light was turned on, the diffraction rings of the 633 nm light was re-established and the light signal could be recorded. The variation of the light signal with the 532 nm light being turned on and off was shown in Fig. 4. In Fig. 4 (a), the light signal could hardly be detected when the 532 nm light was turned off. The signal intensity rose immediately when the 532 nm light was turned on, which was marked by an arrow. When the 532 nm light was turned off again, the signal dropped rapidly. However, the dropped signal intensity exhibited a sudden rise, and then fell back to zero quickly. In addition, the signal intensity kept unchanged for a long time under the irradiation of the 532 nm light as shown in Fig. 4 (b). It indicated that the diffraction pattern of the controlled light was stable over the time under a given power level of the controlling time.

 

Fig. 4. The normalized signal intensity recorded by the photodetector with the 532 nm light being turned on and off (a) in 4 s, and (b) in 1000 s.

Download Full Size | PDF

For the liquid sample, the controlled light could be excited by the controlling light when there was an intersection angle between them [5–7,10]. However, for the azo dye doped liquid crystals sample, the intersection angle between the two lights exhibited tremendous influence on the experiment results of the SXPM. Figure 5 displayed the patterns of the 633 nm light passing through the azo dye doped liquid crystals under the irradiation of the 532 nm light with different powers when there was an intersection angle of 15° between the two lights. The 633 nm light evolved into multiple circular fringes when the 532 nm light irradiated on the liquid crystals sample. The size of the light pattern and the number of the circular fringes increased with the power of the 532 nm light rising. But the shape of the transmitted light pattern was quite different from the experiment results above. The circular fringes were very thin and the number of those fringes far exceeded the number of the diffraction rings of the 532 nm light. For instance, at least 9 circular fringes could be distinguished with the controlling light power of 30 mW, while only two diffraction rings of the 532 nm light were observed. The modulation effect of the 633 nm light by the same light-control-light system in an azo dye solution sample was observed for comparison. The 633 nm light passing through the azo dye solution sample diffracted into multiple rings under the irradiation of the 532 nm light with the power of 48 mW. Both diffraction patterns collapsed accompanied with the thermal convection of the solution. The ring number of the 633 nm light is not more than that of the 532 nm light. Actually, the light pattern in Fig. 5 was closer to that of a parallel light passing through the liquid crystals sample irradiated by 532 nm light as shown in Fig. 6. The 633 nm light was expanded into a parallel light of 30 mm diameter. A series of concentric rings appeared on the screen under the irradiation of the 532 nm light. With the enhancement of the light power, the number of the concentric rings increased gradually, and the number of the concentric rings was much larger than that of the diffraction rings of the 532 nm. It indicated that when there was an intersection angle between the controlling and controlled light, the controlled light passing through the excited azo dye doped liquid crystals sample interfered instead of evolving into the similar diffraction rings to the controlling light. According to the results above, the all-optical modulation of the controlled light in liquid crystals by SXPM method could only be demonstrated successfully when the two beams were merged together.

 

Fig. 5. The photographs of the 633 nm light (10 mW) passing through the azo dye doped liquid crystals under the irradiation of the 532 nm light with the power of 0, 10, 20, 30, 40 and 48 mW. The 633 nm light was focused by a lens (f = 250 mm). The distance between the liquid crystals sample and the white screen was approximately 110 cm. Scale bars, 10 mm.

Download Full Size | PDF

 

Fig. 6. The photographs of the expanded 633 nm light (10 mW) passing through the azo dye doped liquid crystals under the irradiation of the 532 nm light with the power of 0, 10, 20, 30, 40 and 48 mW. The distance between the liquid crystals sample and the white screen was approximately 240 cm. Scale bars, 10 mm.

Download Full Size | PDF

4. Conclusion

In conclusion, the controlled light was modulated by the controlling light with SXPM method in azo dye doped liquid crystals when the two lights were merged together. The diffraction rings of the controlled light were obtained under the irradiation of the controlling light. The diameter of the diffraction pattern of the controlled light increased linearly with the power of the controlling light rising. The thermal convection caused by the heat of the laser beam was suppressed and no collapse effect was observed in the diffraction pattern. The circular symmetry of the diffraction pattern and the linear dependence between the diameter of the pattern and the controlling light power made the shape of the diffraction pattern more predictable, by a light-control-light system based on SXPM. The controlled light presented similar diffraction rings as the controlling light, which could be expected to be applied in all-optical information conversion in the near future.