A novel demonstration of an all-optically controllable dye-doped liquid crystal infiltrated photonic crystal fiber (DDLCIPCF) is presented. Overall spectral transmittance of the DDLCIPCF can decrease and then increase with a concomitant red-shift of the spectrum curve with increasing irradiation time of one UV beam. Continuing irradiation of one green beam following UV illumination on the DDLCIPCF can cause the transmission spectrum to recover completely. The reversible all-optical controllability of the photonic band structure of the fiber is attributable to the isothermal planar nematic (PN) → scattering (S) → isotropic (I) and I → S → PN state transitions of the LCs via the UV-beam-induced trans→cis and green-beam-induced cis→trans back isomerizations of the azo-dye, respectively, in the cladding of the DDLCIPCF. The photoinduced appearance of the S state and the variation of the index modulation between the core and the cladding of the fiber result in the variation of overall spectral transmittance and the shift of transmission spectrum, respectively.
© 2011 OSA
Photonic crystal fiber (PCF) has been widely investigated during the last decade because of its interesting light-guiding mechanisms, many peculiarities (such as high nonlinearity, high numerical aperture, large mode area, and endlessly single mode), and potential applications [1–3]. The features of a PCF are determined by an unusual structure with a spatially periodic microstructure having air array microholes in the silica cladding around a solid or hollow core. Light in a PCF with a solid core of high refractive index is mainly guided via modified total internal reflection similar to a conventional single-mode fiber. Light in a PCF with a hollow core, however, is guided by a so-called photonic bandgap (PBG) effect [1–3]. Such a spatially periodic, microstructured air microhole array not only affects the optical properties of a PCF, but also provides an opportunity to create tunable fiber-based devices by infiltrating high-index materials such as polymer , oil , and liquid crystal (LC) [6–20,25,26] into the air microholes.
Among those high-index materials infiltrated PCFs mentioned above, the LC-infiltrated photonic crystal fiber (LCIPCF) attracts significant attention because its photonic band structure can be easily controlled by changing the LC structure in the microholes, and thus the index modulation between the cladding and the core in a thermal [6–12, 25, 26], electric [9–11, 13–16, 18,20], or optic approach [6, 17]. Alkeskjold et al., in particular, investigated an optically modulated dye-doped LCIPCF (DDLCIPCF) based on the mechanism of a local-heating effect via illumination on the dye. Hsiao et al. reported a light-switchable photoresponsive LCIPCF, in which the transmittance of the PCF can be switched in one second by the isotropic→nematic phase transition of LCs via the cis→trans isomerization of azo-dye. For the first time, a peculiar DDLCIPCF is demonstrated in which photonic band structures can be reversible and all-optically controlled. Overall spectral transmittance in the transmission spectrum of the DDLCIPCF can be controlled to decay and then rise with a concomitant red-shift of the spectrum curve with the increase of irradiation time of one UV beam. Successive irradiation of one green beam on the DDLCIPCF following UV illumination can completely recover the photonic band structure of the DDLCIPCF. The reversible all-optical controllability of the DDLCIPCF is attributable to the isothermal planar nematic (PN) → scattering (S) → isotropic (I) and I → S → PN state transitions of the LCs via the UV-beam-induced trans→cis and green-beam-induced cis→trans back isomerizations of the azo-dye in the cladding, respectively. The photoinduced scattering state, as well as the variation of the index modulation between the core and cladding, results in the variation of overall spectral transmittance and the shift of the transmission spectrum, respectively. The light-induced thermal effect of the dye is excluded from the possible mechanism of the all-optical controllability of the DDLCIPCF.
2. Sample preparation and experimental setups
The mixture used in the present work is dye-doped liquid crystal (DDLC), including 70wt% NLC E7 (ne=1.7391 and no=1.5233 at 25 °C for λ=589 nm; ni≅(ne+2no)/3=1.5952 in the isotropic phase, from Merck) and 30 wt% azo-dye 4MAB (4-Methoxyazobenzene, from Fluka). The uniform mixture is injected into the spatially periodic air microholes of the hexagonal cladding in an all-silica 2cm-length PCF (LMA-20, from NKT Photonics A/S, Denmark) via the capillary effect, forming a DDLCIPCF. Figure 1(a) displays the cross-sectional image of the DDLCIPCF with a 230 μm diameter. The diameters of each microhole in the cladding and the core of the PCF are d=6.4 and D=20 μm, respectively, and the distance between adjacent microholes is 13.2 μm. To identify the orientation of LCs in the microholes of the cladding region, the DDLCIPCF lateral image is observed under a polarizing optical microscope (POM) with crossed polarizers (P⊥A). Figures 1(b) and 1(c) show the transmitted images of the DDLCIPCF as its fiber axis is oriented at ϕ=45° and 90° with respect to the transmission axis of the polarizer, respectively. The DDLCIPCF image in Fig. 1(b) is clearly much brighter than that in Fig. 1(c), implying that the LC director in the microholes is roughly aligned along the fiber axis (planar alignment), thus producing results similar to those presented in previous investigations [6, 19, 20].
Figures 2(a) and 2(b) present two different experimental setups for measuring the cross-sectional image and the transmission spectrum of the DDLCIPCF, respectively, under all-optical control. One white beam from a white light source is guided by a single-mode fiber A and coupled into the input end of the DDLCIPCF by aligning these two fibers on an xyz-stage. One UV beam (with a fixed intensity of 237.5 mW/cm2 and a variable irradiation time tUV) from a UV light source and one green beam (with a variable irradiation time tG) from a diode-pumped solid state (DPSS) laser source (532 nm) are used to sequentially illuminate the DDLCIPCF. The ways to guide the green beam with various intensities to irradiate the DDLCIPCF are different in the two setups. In Fig. 2(a), the signal of the transmitted image of the cross section for the DDLCIPCF at its output end is guided by an objective lens (×10) and then recorded by a CCD camera. The green beam, with 353.7 mW/cm2, is reflected by a nonpolarizing beam-splitter (NBS) and collected by the objective lens and then illuminates the DDLCIPCF. In Fig. 2(b), the transmitted signal at the output end of the DDLCIPCF is coupled into a single-mode fiber B and guided into a spectrometer (USB2000, Ocean Optics) to record the transmission spectrum of the output signal. The green beam with 955.4 mW/cm2 is guided by the single-mode fiber B and then illuminates the DDLCIPCF.
3. Results and discussion
3.1 UV-visible absorption spectra of the azo-dye in a homogeneously-aligned DDLC cell
Before the experimental examination of the all-optical controllability of the DDLCIPCF, auxiliary data, associated with the absorption spectrum in UV-visible region for the 4MAB azo-dye in a homogeneously-aligned DDLC plane cell (length × width × thickness: 3 cm × 2 cm × 7 μm), are pre-acquired. The blue and red curves in Fig. 3 show the measured absorption spectra of the azo-dye before and after irradiating the cell by the UV light with 237.5 mW/cm2 for approximately two minutes, respectively. The 4MAB dye has two absorption bands exhibiting in UV (around 375 nm) and visible (around 445 nm) regions, which are associated with π-π* and n-π* transitions, respectively. The azo-dyes are usually stable themselves at the rod-like trans state in the dark, as displayed in the inset of Fig. 3. With the irradiation of the UV light, massive trans-4MAB molecules can be photoisomerized rapidly into a curved cis state such that the absorption peaks at 375 and 445 nm drops and rises, respectively. Moreover, cis-4MAB can be rapidly transformed into a trans state with the irradiation of a long-wavelength green light or slowly via thermal relaxation. Subsequently, the absorption spectrum of the azo-dye may return to the initial state.
3.2 Transmitted CCD image of the cross section and the normalized transmitted intensity of the all-optically controllable DDLCIPCF
Figure 4(a) shows the variation of the CCD image of the cross section at the output end of the DDLCIPCF while increasing the irradiation time of the UV light (with a fixed intensity of 237.5 mW/cm2) from tUV=0 to 660 s. Output brightness on the DDLCIPCF core decreases from high to low levels by increasing tUV from 0 to 20 s, and then increases to a brighter level by increasing tUV from 20 to 660 s. A summary of the experimental results is seen in Fig. 4(b), in which the variation of the normalized transmitted intensity (defined as the transmitted intensity divided by the strongest transmitted intensity) at the output end of the DDLCIPCF with the tUV is presented. Apparently, experimental results in Figs. 4(a) and 4(b) imply that the LCs in the cladding region of the fiber may go through three different states corresponding to the relative extremities of the transmitted intensity curve during the increase of tUV. The normalized transmitted intensity first decreases from a high to its weakest level from state 1 at tUV=0 s to state 2 at tUV=20 s, and then increases to the strongest level from state 2 to state 3 at tUV=660 s. The mechanism models associated with the three different states will be addressed and discussed in Section 3.3. Following UV irradiation, the DDLCIPCF is irradiated with one green beam [using the setup in Fig. 2(a)] with a fixed intensity of 353.7 mW/cm2 and an increasing irradiation time tG from 0 to 60 s. Associated experimental results are displayed in Figs. 5(a) and 5(b), in which the variations of the CCD image of the cross section and the normalized transmitted intensity at the output end of the DDLCIPCF with increasing tG are acquired, respectively. Either the core brightness or the normalized transmitted intensity at the output end of the DDLCIPCF first decreases from a highest to a weakest level from state 3 at tG=0 s to state 2 at tG=15 s, and increases to a high level from state 2 to state 1 at tG=60 s. Evidently, the changing tendency of the normalized transmitted intensity of the DDLCIPCF, with the increase of tG, is reversed to that with the increase of tUV as shown in Fig. 4. Therefore, the present DDLCIPCF is demonstrated to exhibit an all-optically (reversible) light-guiding controllability under the successive irradiation of one UV and one green beam.
3.3 Mechanism models associated with the all-optical controllability of the DDLCIPCF
Since a separate experiment shows that no optically controllable feature can be found if no azo-dye is added in the LCs infiltrated in the PCF, the azo-dye plays a key role in the previously obtained experimental results (Sec. 3.2) for the all-optical controllability of the DDLCIPCF. As shown in Fig. 6 , the main mechanism for the reversible all-opticalcontrollability of the DDLCIPCF are, accordingly, attributed to the isothermal N→I and I→N phase transitions of the LCs via the UV-beam-induced trans→cis and green-beam-induced cis→trans back isomerizations of the 4MAB dyes . The azo-dyes are generally stable in a rod-like trans state in the dark. The trans-4MAB dyes tend to align along the long axes of the LC molecules due to the guest-host effect in the DDLC. The trans-isomers can absorb UV photons and promptly transform to bent cis-isomers by trans→cis isomerization, and thus disturb the LC host. The concentration of 4MAB dyes converting to cis-state can increase with increasing tUV so that the LC host can gradually change isothermally from N to I phase. While the UV light is turned off and the DDLC is turned to be irradiated by the green light, the cis-4MAB dyes can transform rapidly back to the trans-state via cis→trans back isomerization . The concentration of the cis-4MAB converting back to the trans-state via consecutive cis→trans back isomerization can lead to the reverse phase transition of LCs from I to N as tG increases.
A model of the reversible transformation of the LC state in the cladding of the DDLCIPCF to explain the above experimental results of the all-optically controllable fiber is described as follows. As displayed in Figs. 1(b) and 1(c) [also in Fig. 8(a) on the next page], the LCs in state 1 in the irradiated region of the DDLCIPCF is confirmed to lie in N phase with a planar configuration (PN state). A model describing the PN structure in the cladding region of the DDLCIPCF is illustrated in Fig. 7(a) , in which the LC molecules tend to align themselves along the fiber’s axis. As indicated in Figs. 8(b) and 8(c), the observed lateral transmitted images of the DDLCIPCF, under a POM with same configurations shown in Fig. 1 (P⊥A with ϕ=45° and 90°), can further identify the LC states of the cladding region, the states 2 and 3, respectively. The POM images in the irradiated region of the DDLCIPCF in Fig. 8(c) in the two configurations both present as dark figures, as the LCs in the cladding lies in state 3 (at tUV=660 s or tG=0 s). The results imply that the overall LCs in state 3 has lost its birefringence, and thus is in an isotropic phase in the whole irradiated region of the fiber. Figure 7(c) presents the model of the isotropic LC structure in state 3 within the cladding region due to the isothermal N→I phase transition of the overall LCs via UV-beam-irradiation-induced trans→cis isomerization of the azo-dyes in the whole irradiated region. Figure 8(b) shows peculiar POM images of the DDLCIPCF, while the LC lies in state 2 under the irradiation of the UV beam (at tUV=20 s) or green beam (at tG=15 s). The magnified images of the DDLCIPCF [the insets of Fig. 8(b)] reveal a multi-domain-like LC structure with the coexistence of the PN and I states in the bright and dark regions. While the incident white beam propagates inside the DDLCIPCF, the multi-domain-like LC texture in the cladding can cause a strong scattering effect on the propagating beam, resulting in the significant decay of the normalized transmitted intensity of the incident light in state 2 (as shown in Fig. 4). The model of the multi-domain-like LC texture in state 2, within the cladding, is plotted in Fig. 7(b), in which the PN and I states of the LCs can coexist in the cladding’s microholes. The scattering state was already observed in Tabiryan’s previous investigation based on a photosensitive azobenzene LC plane cell . A separate experiment also observes a similar scattering effect using a plane cell with the same DDLC materials under UV irradiation (associated result is not shown herein). The presentation of such a scattering state can be primarily attributed to a nonhomogeneous phase transition of the LCs caused by a nonhomogeneous concentration distribution of cis-isomers subjected to nonhomogenous light on the DDLC sample. Therefore, when tUV increases, both the number and size of the nucleus with N→I phase transition in the fiber microholes increase, resulting in increased scattering. However, scattering will decay once the isotropic nuclei overgrow and merge with each other under continuous UV irradiation. In the final stage, LCs at the entire UV-irradiated region becomes isotropic, and the scattering vanishes. Upon switching off the UV beam, the successive irradiation of the green beam with increasing tG can lead to an
overall increase of the concentration of the trans-isomers via cis→trans back isomerization, which can result in a variation of LC texture in the irradiated region. This variation is reversed to that caused by UV irradiation with increasing tUV, which is the reason that the LCs in the cladding of the DDLCIPCF undergo reversible PN→S→I and I→S→ PN state transitions under the successive irradiation of the UV and green beams.
3.4 Transmission spectrum of the all-optically controllable DDLCIPCF
To understand the all-optical controllability for the photonic band structure of the DDLCIPCF, the transmission spectrum for the output signal of the incident white beam is further measured via the DDLCIPCF under successive irradiation of the UV beam (237.5 mW/cm2) and the green beam (955.4 mW/cm2) based on the setup in Fig. 2(b). Figures 9(a) and 9(b) show the variations of transmission spectrum of the incident white beam in the spectral region of 500–750 nm through the DDLCIPCF, with increasing tUV from 0 to 20 s and from 20 to 660 s, respectively. The transmission in the vertical axis is defined by 10 × log (It/I0) (dB), where It and I0 represent the transmitted and incident intensities of the white beam, respectively. The mechanism to guide the white beam in the DDLCIPCF is clearly the PBG effect because of the appearance of the series of bandgaps in the transmission spectrum.
As displayed in Fig. 9, not only the overall spectral transmittance, but also the spectral position of the series of the bandgap for the DDLCIPCF, can be modulated by increasing tUV from 0 to 20 s and then from 20 to 660 s. The reason for the modulation of the overall spectral transmittance will be explained later. The bandgap series in the transmission spectrum presents a continuous red-shift as tUV increases from 0 to 660 s. The red-shift of the bandgaps for the fibers in the visible region can be explained qualitatively based on a simple analytical Antiresonant Reflecting Optical Waveguide (ARROW) model . Each microhole of the cladding with LCs can be approximately regarded as isolated waveguides if the infiltrated LC index in each microhole is higher than that of silica in the surrounding. The minima in the transmission spectrum of the PCF can occur at model cut-off wavelengths for the optical modes supported by a single isolated LC-filled waveguide. According to the ARROW model, the transmission maxima result from antiresonant wavelengths that undergo destructive interference within the high-index cladding region so that light can be confined in the low-index core. The wavelengths λm for the minima of the transmission spectrum in the isotropic case can be given by the following formula for λ << D [23, 24, 6, 13]6, 13], the above model cut-off approach can be suitable for use in short-wavelength regions (e.g., the visible region in the present case) for the planar LCIPCF. Since the LCs in the cladding of the DDLCIPCF is in the PN state, the LC index experienced by the incident beam propagating roughly along the fiber axis is primarily the ordinary one no. As known from the previous section, the value of no of LCs in the visible region can be modulated to increase from 1.5233 to 1.5952 (for λ=589 nm) when the LCs changes from PN to I state with increasing tUV. Taking the increasing values of LC indices (from no=1.5233 to ni=1.5952) into n2 in Eq. (1), we can obtain increasing corresponding values of wavelength λm for the minima of the transmission spectrum, whose result is qualitatively in agreement with the experimental data that the series of the bandgap in the transmission spectrum of the DDLCIPCF red-shifts as the tUV increases from 0 to 660 s (Fig. 9). Moreover, the ARROW model concludes that the wavelengths in Eq. (1) give the cut-off wavelengths of the individual LC hole and must coincide with the minima of the transmission spectrum of the DDLCIPCF. The red-dotted vertical lines marked in Fig. 9(a) [Fig. 9(b)] represent the calculated cut-off wavelengths of the optical mode guided in an isolated LC hole at PN (I) state based on Eq. (1), where the modal index n2(λ) of the optical modes in the LC hole at PN and I states are estimated by the available data of the wavelength-dependent index of E7 at the N and I phases [no(λ) and ni(λ), respectively] in Ref. 6. We can see that the experimental data are consistent to the theoretical prediction of ARROW model. Notably, the ARROW model is not suitable to be used within the wavelength region shorter than 550 nm because of the strong absorption of the azo dye.
The reason for the optical controllability of the overall spectral transmittance of the DDLCIPCF under the UV irradiation presented in Fig. 9 is explained as follows. The overall spectral transmittance of the fiber first decreases at tUV=0→20 s and then increases at tUV=20→660 s [Figs. 9(a) and 9(b), respectively]. The variation of overall spectral transmittance is consistent with data obtained in the variations of the CCD images of the cross section of the DDLCIPCF and of its normalized transmitted intensity [Figs. 4(a) and 4(b), respectively]. As explained in Section 3.3, under the irradiation of the UV beam with increasing tUV at 0→20 s, the multi-domain-like LC texture in the cladding region will appear to cause an S state, resulting in the significant decrease in overall spectral transmittance of DDLCIPCF. While tUV continually increases from 20 to 660 s, the isotropic LC nuclei overextend and merge with each other, such that the scattering decays in the cladding region, leading to the increase of the overall spectral transmittance of the fiber. At tUV ≥ 660 s, the overall LCs in the irradiated cladding region becomes isotropic such that the overall spectral transmittance achieves a significantly high level. Note that the transmittance at the positions of both the maxima and minima of the spectrum increases at tUV=20→660 s in Fig. 9(b). The transmission bands in the transmission spectra are mainly contributed from the guiding light at the region of the fiber core, which is identified by comparing CCD images in Fig. 4(a) and the transmission spectra in Fig. 9(a) at tUV=0→20 s. At tUV=20→660 s, the increase of brightness in the cladding region of the fiber, displayed in Fig. 4(a), causes the increase of transmittance in the positions of the minima in the spectrum. In this interval, the increase of the overall transmittance in the spectrum can be attributed to the decrease of scattering loss from the anisotropic to isotropic orientation of LCs in the DDLCIPCF [12, 25].
After irradiating the DDLCIPCF for 660 s, the UV beam is turned off and simultaneously the green beam with 955.4 mW/cm2 is turned on to irradiate the fiber at tG=0, where tG represents the irradiation time of the green beam. Figures 10(a) and 10(b) show the variations of the transmission spectrum of the fiber with increasing tG=0→10 s and 10→50 s, respectively. The overall spectral transmittance decreases at tG=0→10 s, and then increases at tG=10→50 s, with a concomitant blue-shift in the curve of the transmission spectrum. Evidently, the influence of the green-beam-irradiation with increasing tG on the photonic band structure of the fiber is reversal of that of the previous UV-beam-irradiation with increasing tUV. As mentioned in Section 3.3, the overall concentration of the cis-4MAB in the cladding decreases while that of the tran-4MAB increases with increasing tG via green-beam-induced cis→trans back isomerization. The disturbance of the cis-isomers on the LCs decreases and the LCs may change from I back to PN states. The number and size of LC nucleus with I→PN state transition both increase with increasing tG=0 to 10 s, such that the scattering from the multi-domain-like LC texture increases, resulting in the decrease of overall spectral transmittance of the DDLCIPCF [Fig. 10(a)]. At tG=10→50 s, the over-expanding LC regions with PN state merge with each other to induce the decay of the scattering, and thus the rise of the overall spectral transmittance of the fiber [Fig. 10(b)]. The concomitant blue-shift of the transmission spectrum at tG=0→50 s is just a reversal of the red-shift of the transmission spectrum at tUV=0→660 s. Since the LC index experienced by the propagating wave in the fiber gradually decreases from ni to no at tG=0→50 s, the series of the bandgap in the transmission spectrum blue-shifts, which can also be qualitatively confirmed by Eq. (2). Low transmittance in the green region of the transmission spectrum of the output of the DDLCIPCF results from the strong absorption of the azo-dye.
3.5 Exclusion of thermal-effect-induced phase transition for the mechanism of the all-optical controllability of the DDLCIPCF
The following two separate experiments will elucidate that the all-optical controllability of the DDLCIPCF is not attributable to the thermal-effect-induced phase transition of LCs in the cladding region under the photo-excitation of the azo-dye. As mentioned in Section 3.3,the LCs in the cladding lies in isotropic state once the fiber is irradiated by the UV beam with 237.5 mW/cm2 for 660 s. At the moment tUV=660 s, the UV beam is turned off. The state of the LCs in the cladding can then very slowly recover from I to S and then PN states due to the thermal cis→trans back isomerization of the azo-dye. Figure 11 shows the variations of the normalized transmitted intensity and corresponding CCD image of the cross section of the DDLCIPCF with the time via such a state transition of LCs. Clearly, the time to induce the I→S→PN state transition of LCs in the cladding of the fiber via the thermal cis→trans back isomerization is as long as tC→T,Δ≅93 hrs which is much longer than that via the green-beam-irradiation-induced cis→trans back isomerization (60 s). In a separate experiment, an un-irradiated DDLCIPCF is placed on the sample platform of the POM in a dark room for observing the state variation of the LCs. The fiber is pre-heated by a vacuum thermal stage (TS102V-STC20A, INSTEC) such that the overall LCs in the cladding become isotropic, and thereupon, the thermal stage is removed. The measured thermal relaxation time (tTR) for the LCs to undergo I→S→PN state transition is less than 10 minutes. The time tTR (<10 min) is much less than tC→T,Δ (93 hr) obtained above, and thus the all-optical controllability of the fiber described previously is not attributed to the photoinduced thermal-effect, but to the isothermal photoisomerization-induced state transition of the LCs in the cladding of the fiber.
The present study is a novel investigation of an all-optically controllable DDLCIPCF. The overall spectral transmittance in the DDLCIPCF transmission spectrum can decrease and then increase with a concomitant red-shift of the spectral bands with the increase of the irradiation time of one UV beam. Subsequent irradiation of one green beam following the UV irradiation on the fiber can entirely recover the transmission spectrum. The reversible all-optical controllability of the photonic band structure of the fiber results from the isothermal PN →S → I and I → S → PN state transitions of the LCs via UV-beam-induced trans→cis and green-beam-induced cis→trans back isomerizations of the azo-dye in the cladding, respectively. The presentation of the S state and the variation of the index modulation between core and cladding induce the variation of the overall spectral transmittance and the shift of the spectral bands of the transmission spectrum of the fiber, respectively. Such an all-optically controllable DDLCIPCF may potentially be applied to the fields of integrated photonics and optical communication devices. Further investigations for the all-optically controllable DDLCIPCF in the near-infrared region in experiments and simulation of related band structure of the fiber will be carried out in the near future.
The authors would like to thank the National Science Council of the Republic of China, Taiwan (Contract number: NSC 97-2112-M-006-013-MY3) and the Advanced Optoelectronic Technology Center, National Cheng Kung University, under projects from the Ministry of Education, for financially supporting this research. We greatly appreciate KGSupport for the editorial assistance.
References and links
5. R. T. Bise, R. S. Windeler, K. S. Kranz, C. Kerbage, B. J. Eggleton, and D. J. Trevor, “Tunable photonic band gap fiber,” in OSA Trends in Optics and Photonics (TOPS) 70, Optical Fiber Communication Conference Technical Digest, Postconference Edition (Optical Society of America, Washington, DC, 2002), 466–468.
6. T. T. Alkeskjold, J. Laegsgaard, A. Bjarklev, D. S. Hermann, A. Anawati, J. Broeng, J. Li, and S. T. Wu, “All-optical modulation in dye-doped nematic liquid crystal photonic bandgap fibers,” Opt. Express 12(24), 5857–5871 (2004). [CrossRef] [PubMed]
7. D. Noordegraaf, L. Scolari, J. Laegsgaard, T. Tanggaard Alkeskjold, G. Tartarini, E. Borelli, P. Bassi, J. Li, and S. T. Wu, “Avoided-crossing-based liquid-crystal photonic-bandgap notch filter,” Opt. Lett. 33(9), 986–988 (2008). [CrossRef] [PubMed]
9. L. Wei, L. Eskildsen, J. Weirich, L. Scolari, T. T. Alkeskjold, and A. Bjarklev, “Continuously tunable all-in-fiber devices based on thermal and electrical control of negative dielectric anisotropy liquid crystal photonic bandgap fibers,” Appl. Opt. 48(3), 497–503 (2009). [CrossRef] [PubMed]
10. T. R. Wolínski, K. Szaniawska, S. Ertman, P. Lesiak, A. W. Domanski, R. Dabrowski, E. Nowinowski-Kruszelnicki, and J. Wojcik, “Influence of temperature and electrical fields on propagation properties of photonic liquid-crystal fibres,” Meas. Sci. Technol. 17(5), 985–991 (2006). [CrossRef]
11. T. R. Wolínski, S. Ertman, A. Czapla, P. Lesiak, K. Nowecka, A. W. Domanski, E. Nowinowski-Kruszelnicki, R. Dabrowski, and J. Wojcik, “Polarization effects in photonic liquid crystal fibers,” Meas. Sci. Technol. 18(10), 3061–3069 (2007). [CrossRef]
12. J. Du, Y. Liu, Z. Wang, B. Zou, B. Liu, and X. Dong, “Liquid crystal photonic bandgap fiber: different bandgap transmissions at different temperature ranges,” Appl. Opt. 47(29), 5321–5324 (2008). [CrossRef] [PubMed]
13. L. Scolari, T. T. Alkeskjold, J. Riishede, A. Bjarklev, D. S. Hermann, A. Anawati, M. Nielsen, and P. Bassi, “Continuously tunable devices based on electrical control of dual-frequency liquid crystal filled photonic bandgap fibers,” Opt. Express 13(19), 7483–7496 (2005). [CrossRef] [PubMed]
14. D. Noordegraaf, L. Scolari, J. Laegsgaard, L. Rindorf, and T. T. Alkeskjold, “Electrically and mechanically induced long period gratings in liquid crystal photonic bandgap fibers,” Opt. Express 15(13), 7901–7912 (2007). [CrossRef] [PubMed]
16. A. Lorenz, H.-S. Kitzerow, A. Schwuchow, J. Kobelke, and H. Bartelt, “Photonic crystal fiber with a dual-frequency addressable liquid crystal: behavior in the visible wavelength range,” Opt. Express 16(23), 19375–19381 (2008). [CrossRef]
17. V. K. S. Hsiao and C.-Y. Ko, “Light-controllable photoresponsive liquid-crystal photonic crystal fiber,” Opt. Express 16(17), 12670–12676 (2008). [PubMed]
18. C.-H. Lee, C.-H. Chen, C.-L. Kao, C.-P. Yu, S.-M. Yeh, W.-H. Cheng, and T.-H. Lin, “Photo and electrical tunable effects in photonic liquid crystal fiber,” Opt. Express 18(3), 2814–2821 (2010). [CrossRef] [PubMed]
19. V. V. Presnyakov, Z. J. Liu, and V. G. Chigrinov, “Infiltration of photonic crystal fiber with liquid crystals,” Proc. SPIE 6017, 60170J, 60170J-7 (2005). [CrossRef]
20. L. Scolari, S. Gauza, H. Xianyu, L. Zhai, L. Eskildsen, T. T. Alkeskjold, S.-T. Wu, and A. Bjarklev, “Frequency tunability of solid-core photonic crystal fibers filled with nanoparticle-doped liquid crystals,” Opt. Express 17(5), 3754–3764 (2009). [CrossRef] [PubMed]
21. H.-K. Lee, A. Kanazawa, T. Shiono, T. Ikeda, T. Fujisawa, M. Aizawa, and B. Lee, “All-optically controllable polymer/liquid crystal composite films containing the azobenzene liquid crystal,” Chem. Mater. 10(5), 1402–1407 (1998). [CrossRef]
22. N. V. Tabiryan, S. V. Serak, and V. A. Grozhik, “Photoinduced critical opalescence and reversible all-optical switching in photosensitive liquid crystals,” J. Opt. Soc. Am. B 20(3), 538–544 (2003). [CrossRef]
23. A. K. Abeeluck, N. M. Litchinitser, C. Headley, and B. J. Eggleton, “Analysis of spectral characteristics of photonic bandgap waveguides,” Opt. Express 10(23), 1320–1333 (2002). [PubMed]
24. N. M. Litchinitser, S. C. Dunn, P. E. Steinvurzel, B. J. Eggleton, T. P. White, R. C. McPhedran, and C. M. de Sterke, “Application of an ARROW model for designing tunable photonic devices,” Opt. Express 12(8), 1540–1550 (2004). [CrossRef] [PubMed]
25. K. Szaniawska, T. R. Wolinski, S. Ertman, P. Lesiak, A. W. Domanski, R. Dabrowski, E. Nowinowski-Kruszelnicki, and J. Wojcik, “Temperature tuning in photonic liquid crystal fibers,” Proc. SPIE 5947, 594705, 594705-6 (2005). [CrossRef]
26. J. Tuominen, H. Hoffrén, and H. Ludvigsen, “All-optical switch based on liquid-crystal infiltrated photonic bandgap fiber in transverse configuration,” JEOS:RP 2, 07016 (2007). [CrossRef]