1. Introduction

Photonics has been an emerging field since the first laser device was invented in the middle of the 20th century. Among the variety of studies, organic laser devices have increasingly attracted a lot of interest because of their flexibility and capability for downsizing [1]. Dynamic manipulation of laser emission frequency can be achieved in such systems by applying a certain stimulus. Tunability is very important in telecommunication applications, absorption spectroscopy, fluorescence excitation, and even time-resolved in-situ studies of biological systems. It can be also interesting for “lab-on-a-chip” applications where all of the necessary processes, including separation of chemicals and characterization of the target, are done very quickly on a small chip [2].

Liquid crystal elastomers and gels are interesting materials with unusual properties, because they combine mechanical characteristics of polymers with variable optical birefringence of liquid crystal phases [3]. Liquid crystal elastomers show a complicated response to mechanical strain, since the conformational changes of polymer chains affect the liquid crystal order: thus they can be used for converting mechanical energy into optical response and vice versa. Their applications can range from artificial muscles, electro-optical switches, electro- or photo-controllable micro-actuators, switchable color-tunable reflectors, full-color reflective displays, and tunable low-threshold mirror-less lasers [4, 5].

 

Fig. 1. A schematic view of CLC elastomer under uniaxial strain imposed along x. Contraction of z dimension, λ_zz, leads to a continuous shift of the photonic bandgap.

Download Full Size | PDF

In the case of cholesteric liquid crystal (CLC) elastomers, mechanical deformation directly affects the selective reflection band, the unique optical property of the cholesteric phase. The opto-mechanical response of CLC elastomers has been studied theoretically [6, 7] and experimentally [8, 9]. When the CLC elastomer is uniaxially stretched by a factor λ along the x direction, perpendicularly to the helical axis z (Fig. 1), the director angle ϕ in the xy plane is given by the equation:

where r is the chain anisotropy parameter [3] given by the ratio of the effective step lengths of a random walk, representing the motion of the polymer chain in the directions parallel and perpendicular to the director, r=l _‖/l _⊥. In the materials used in this work the value of r is approximately 1.2 [8]. Equation (1) has been derived under the assumption of material incompressibility [6]. In a cholesteric elastomer, uniaxially stretched perpendicular to its helical axis, the original theory predicted the anisotropic lateral contraction: λ_zz=λ ^-1/4 and λ_yy=λ ^-3/4; this was later modified by accounting for more delicate interactions [7], predicting λ_zz=λ ^-2/7 and λ_yy=λ ^-5/7. Importantly, the wavenumber of periodic modulation is given by q̃=q ₀/λ_zz, so the position of the selective reflection bands can be continuously blue-shifted by mechanical deformation.

Figure 2 shows how the uniaxial extension λ_xx modifies the director angle ϕ, according to Eq. (1). Without strain, the director follows a simple helix such that ϕ increases linearly from 0 to π along the helix phase q̃z. This structure changes when strain λ is imposed, as the director tends to follow the uniaxial elongation. When the strain exceeds a certain threshold value, the deformed helix experiences a transition (shown by the arrow in Fig. 2), in which the middle point (q̃z=π/2) suddenly changes from π/2 to 0, yielding to point along the stretching direction x. At this point the system loses its original helicity, forming a periodic structure having no phase chirality.

As mentioned earlier, the cholesteric phase selectively reflects light whose wavelength matches its helix periodicity. Many studies have been carried out in order to utilize the characteristic gap in the photonic spectrum for lasers, and the principles of lasing mechanism is reviewed by Kopp et al in [10]. What makes CLC unique for a versatile organic laser system is that the emission resonance is expected to occur without the use of conventional mirrors. This is a consequence of the natural distributed feedback structure of cholesteric helix where the photonic bandgap is “felt” everywhere inside the medium. Similarly to a layered structure alternating different refractive indices, monodomain cholesteric phase acts as a one-dimensional photonic crystal. Inside the gap electromagnetic waves cannot propagate, and the density of states is maximized at the band edge. In cholesterics liquid crystals the periodic structure has a chirality so that it is only effective when the handedness of circularly polarized light matches the cholesteric helicity. However, in the deformed chirality-free structure above the strain threshold there is no longer any chiral selectivity and all light of the appropriate wavelength is restricted everywhere in the material.

 

Fig. 2. Director angle ϕ against position along the pitch z, for a system with anisotropy r=1.2. Without stress, at λ=1, ϕ increases linearly with the position q̃z. As λ increases, the curvature changes until the point when the system undergoes a chirality-loss transition (shown by a arrow), at which the original helical structure becomes a non-chiral periodic structure.

Download Full Size | PDF

Since the lasing frequency is locked to the position of the photonic band gap, the emission can be tuned by changing the helix pitch by any stimuli or modifications of the outer environment. It has been shown that the position of the bandgap, and therefore the tunability of lasers, can be achieved in many ways: by creating a gradient in chiral dopant concentration by changing the solubility with a temperature gradient [11, 12], by applying an electric field [13], by creating a liquid crystal cell which contains several dyes [14, 15], by adding a UV-sensitive chiral dopant which changes the pitch of the helix with photoisomerization [15, 16, 17, 18, 19, 20], by disrupting the hydrogen bonds of a chiral phase with the exposure to charged aminoacids [21] or by applying mechanical stress in cholesteric elastomers [22, 23, 24].

The aim of the first part of this work is to investigate how crosslinking affects the spectral properties of CLC elastomers and the tunability of the photonic bandgap by mechanical deformation. Mechanical tunability of the photonic spectrum is studied under uniaxial stretching and compared to the theoretical model [6].

In the second part we investigate the lasing tunability of a system composed by an isotropic layer containing Rhodamine B dye “sandwiched” between two cholesteric elastomers: the data suggest that the cholesteric elastomers act as distributed feedback mirrors and that this system allows good mechanical tunability and reasonably low lasing threshold.

2. Materials and Methods

2.1. Sample Preparation

Figure 3(a) shows the composition and the components of our cholesteric (CLC) elastomer. The classical polysiloxane side-chain system [25, 26] was chosen to produce material with a low glass transition temperature, to allow high mechanical deformations at room temperature. The side-groups attached to the polymer backbone via hydrosilation reaction were 4-pentylphenyl-4’-(4-buteneoxy) benzoate, cholesterol pentenoate, and 3-methyl-3-oxetane-(11-undecenemethenether), playing the roles of the mesogenic unit, the chiral dopant, and the UV-activated crosslinker, respectively. The mesogenic groups were used in previous works, where details on preparation can be found [8, 31, 27]. The new UV-crosslinker was chosen since the oxetane structure undergoes a rapid cationic ring-opening reaction under the presence of a proper initiator [28], and does not produce side-products after the reaction. The resulting polymer has an isotropic phase above T_c=73°C, the cholesteric phase below T_c, and the glass transition at T_g≈9°C.

 

Fig. 3. (a) Components and the reaction scheme of UV-crosslinkable choleseric polysiloxane; (b) Components and the reaction scheme of the dye-containing modified side-chain polysiloxane.

Download Full Size | PDF

In addition to UV-crosslinkable CLC polymers, we have also prepared modified side-chain (MS) polysiloxane polymers, designed as a medium to homogeneously dissolve and contain a variety of hydrophobic laser dyes. Figure 3(b) shows the composition of this polymer. The laser dye used in this work was Rhodamine B (from Aldrich) due to its convenient spectral absorption position with respect to the cholesteric pitch at a given chiral dopant concentration.

After synthesis, all side-chain polysiloxane polymers (CLC or MS) were dissolved in acetone and filtered to remove all insoluble components. Following filtration, a 1 wt% of cationic UV initiator 1-Butyl-3-methylimidazolium chloride was dissolved in the solution. For MS polymers, 0.5wt% of the laser dye was dissolved, and then solutions were vacuum dried at 40°C overnight. In order to protect UV crosslinkable groups, all of the processes were carried out in the dark. A basic silicone elastomer (SYLGARD 184, from Dow Corning) was used as a mechanical support for the thin films of active materials.

For optical measurements, a well aligned monodomain cholesteric film was required. The alignment was achieved by sandwiching a CLC polymer between two quartz plates, the surface of which was coated with a water-soluble uniaxially rubbed sacrificial layer of poly(vinyl) alcohol (PVA). The two plates were positioned in an antiparallel rubbing orientation. Two 10µm spacers were inserted between the plates in order to control sample thickness. The assembled liquid crystal cell was kept at 90°C (in the isotropic phase) for more than three hours in order to cancel any mechanical hysteresis in polymer melts, then slowly cooled below the transition temperature (to 60°C) and kept overnight for alignment. Finally, after cooling down to room temperature, the liquid-crystal polymer cell was UV-irradiated for three minutes on both sides in order to cross-link the polymers. In some cases we did not add the photo-initiator to the solution, therefore we did not irradiate the CLC films. Although some occasional reaction between the cross-linking groups can never be fully avoided, we were satisfied that no elastomer network was formed. After the full treatment, the cells were soaked in water at 4°C and kept until all of the components were disassembled by PVA layer dissolution (approximately a week). The film was then collected on a silicone elastomer strip in cold water (to keep the film in a glassy state where it was easier to handle), then dried at room temperature and in the vacuum at 40 °C.

 

Fig. 4. Scheme of the three layer sandwiched configuration.

Download Full Size | PDF

MS polymer films were prepared similarly to the CLC polymer films. For a laser experiment, CLC-MS-CLC sandwich structure was assembled as shown in Fig. 4. A MS layer was placed on top of a CLC polymer film which was mounted on a silicone elastomer strip as previously mentioned. The other side of MS layer was then treated similarly by another CLC polymer film on a silicone elastomer strip.

2.2. Measurements

Single-layer CLC films on silicone elastomer support, or assembled CLC-MS-CLC sandwich systems, were stretched by using a mechanical stretching device with a micrometric screw, symmetrically from both ends in order to minimize the displacement of the measuring spot.

Transmittance spectra of stretched CLC films were measured in order to study the effect of sample uniaxial elongation on the selective reflection wavelength. For transmission spectra, a quartz lamp equipped with a constant-current power supply was used as a light source. The intensity of light was optimized by an adjustable slit positioned at the window of the light source. The incident light was circularly polarized by a combination of a linear polarizer and a Fresnel rhomb [8]. The linear polarizer was set to either +45° or -45° of the optical axis to yield right and left handedness of incident light, respectively. The light was normally incident on the sample mounted in the stretching device. The transmitted light was guided by an optical fibre to a spectrometer (Thermo-Oriel MS260i) with a spectral resolution of 1.5nm, equipped with a CCD camera. The transmittance data was collected every 0.021 seconds, then a mean of a hundred measurements was recorded. All measurements were conducted at room temperature in the ambient light conditions.

Data analysis of spectra was carried out by a Matlab program consisting of three subroutines: Peakdetect.m, deriv.m, and fastsmooth.m, available on www.mathworks.com. After the baseline was subtracted from the original data, a peak height, a peak wavelength, and a peak width at the half height were calculated. In the measurement of spectrum evolution in time after instantaneously imposed deformation, all of the baselines of the individual spectra were normalized in order to obtain a common baseline.

Lasing emission of the CLC-MS-CLC three layered structure and its mechanical tunability by elongation were investigated under pumping with a 532nm pulsed laser beam (pulse duration≈500ps) from a frequency-doubled Nd:YVO4 laser (AOT). The intensity was adjusted by a filter. The light was left-hand circularly polarized by a combination of a linear polarizer and a Fresnel rhomb. The left-handedness of pulsed laser was used as a reference minimum reflection from a CLC polymer. The pumping laser beam was focused on the sample surface at 35° incidence and the emission from the sample was detected by a spectrometer (USB 4000; Ocean optics), positioned perpendicularly to the sample surface, with a spectral resolution of less than 1nm. Beam spot size on the sample surface was calculated by the following method: a sample holder was replaced by a blade fixed on a movable stage equipped with a micrometric position controller. Beam intensity behind the blade was then measured by a power meter (HOVAII and OPHIR) and data was collected with different blade positions; fitting these data points with a gaussian distributions provides the diameter w of the beam spot [29]. The energy density of pumping source was calculated by dividing the pumping energy by the beam spot area.

 

Fig. 5. Transmittance spectra of non-crosslinked and crosslinked CLC films under uniaxial strain, under right- and left handed circularly polarized incident light (R* and L* plots, respectively). L* spectra show a pronounced selective reflection band in the stretched crosslinked CLC elastomer, whereas no L* peak is seen in the non cross-linked CLC polymer.

Download Full Size | PDF

3. Results and Discussion

Figure 5 shows transmittance spectra of cross-linked and non cross-linked CLC polymer under a given uniaxial strain. Figure 6 shows the peak height of each spectrum as a function of the sample elongation. The incident light was circularly polarized in order to study evolution of the non-chiral periodic structure induced by mechanical deformation [6].

In the non-crosslinked CLC polymer film, the selective reflection was only observed for the right-handed circular polarization of incident light. The puzzling observation of a significant R* peak shift, indicating a change in helix pitch, even in the non-crosslinked system, remains a challenge for us to explain. This may be a kinetic effect: even without any crosslinking, the CLC polymer is very viscous below T_c and the chains are highly entangled; these dynamic constraints force the helix to accommodate the shape change by changing the helix pitch accordingly. Relaxation (time-dependent) studies were conducted in this work, but we were not able to achieve a full return of the helical pitch to its original (undeformed) value of 765nm even after a full day of relaxation.

 

Fig. 6. Peak height of the selective reflection band against uniaxial elongation. On stretching, the peak height of L* polarization in crosslinked CLC elastomers becomes comparable with the R* polarization reflection, showing that the original helical chirality is lost in favor of a non-chiral periodical texture. No significant change is seen in the non-crosslinked CLC polymer reflection intensity. The lines are just a guide for the eye.

Download Full Size | PDF

In the crosslinked CLC elastomer network the evolution of the birefringent texture on deformation broadly follows the theoretical predictions and earlier observations. The selective reflection band in the left-handed incident light (opposite to the initial helical handedness) appears and becomes comparable with the one in the right-handed incident light at above 30% elongation. This proves that the original helical director structure loses its chirality and becomes a periodic achiral stack of alternating layers (see Fig. 1). From the comparison between non-crosslinked and crosslinked CLC systems, it is also clear that the local anchoring effect of crosslinking is essential for the director response to deformation. One can see in Fig. 5 that the overall transmittance is lower in the crosslinked elastomer than in the non-crosslinked CLC polymer. This indicates that the sample has a higher degree of disorder, represented in a small but unavoidable variation in the helical pitch value and direction: the disorder is a consequence of the fact that crosslinkers are not mesogenic units and thus they introduce small local quenched distortions in the order parameter field.

As shown in Fig. 1, a uniaxial stretching in the x direction (perpendicular to the helical axis) results in contraction in the y direction and the z direction (parallel to the helical axis). The contraction of the helical pitch, and hence the contraction of the system in the z direction (strain λ_zz, assumed affine with the pitch) is calculated from the peak shift. Figure 7 plots λ_zz against the imposed elongation λ_xx to demonstrate how the system is deformed by the uniaxial stretching. In the case of non-crosslinked CLC polymer, λ_zz follows the power law λ_zz=λ ^-0.5 _xx up to very high extensions λ_xx=2 and above. In this state, λ_yy also follow the same power law, λ_yy=λ ^-0.5 _xx, because of the overall incompressibility of the system demanding λ_xxλ_yyλ_zz=1. Therefore, the mechanical contraction occurs isotropically in both y and z direction in spite of the symmetry breaking due to the cholesteric helix in the z-direction.

On the other hand, in the case of crosslinked CLC elastomer, the pitch contraction λ_zz is much slower. Fitting in Fig. 7 gives λ_zz=λ ^-0.38 _xx, which means that λ_yy=λ ^-0.62 _xx. Therefore, the contraction occurs anisotropically. Although this is an expected phenomenon, as mentioned in the introduction, the value of scaling exponent is smaller than the theoretical value of -1/4 [6] (for comparison, that line is drawn on the plot). Possible reason could be that our system has a much lower cross-linking density than that considered in the theory and the local anchoring effect for weakly crosslinked material was not enough to make the system ideal. This weaker anchoring effect is also suggested by the fact that the critical strain, where the helicity is lost, is ≈40%, bigger than the theoretical value of ≈20%. Nevertheless, some deviation from the fitted power law is observed around the critical strain, as the theoretical models also suggest.

 

Fig. 7. The contraction of cholesteric helix λ_zz against imposed uniaxial strain λ_xx. The non cross-linked CLC polymer plots on the line of λ ^-0.5 _xx showing the isotropic deformation, whereas the cross-linked CLC polymer plots on the line of λ ^-0.38 _xx showing the anisotropic deformation.

Download Full Size | PDF

In order to investigate the dynamics of cholesteric pitch evolution under strain, the relaxation of the helix under constant strain was studied by measuring transmittance spectrum evolution with time, after imposed deformation. The normalized peak shift of the selective reflection band Δλ (t) was defined by the relation:

where λ (I) is the initial peak wavelength without strain, λ (t) is the measured peak wavelength at a given time, and λ_min(0) is the minimum peak wavelength (maximal shift) obtained at time t=0. By this definition, Δλ(t) stays equal to unity when the helix maintains the most contracted state without any relaxation, and it becomes zero if the pitch can return back to the initial non-strained value after a long relaxation. Figure 8 plots the evolution Δλ(t) against time. The samples used here were non-crosslinked CLC polymer, the crosslinked CLC elastomer (the same as in Fig. 7), but also a very weakly crosslinked CLC polymer, where the crosslinking density was not enough to show the selective reflection band in the left-handed circularly polarized light. The samples were instantaneously stretched by 70% through the silicone support. Obtained data are fitted with high accuracy by a double exponential equation:

where k_s and k_f are slow and fast kinetic rates, respectively. Remarkably, the fitted value for the fast-rate constant is the same in all three materials studied, k_f=1.0×10^-1 s ^-1. In contrast, the slow relaxation rate strongly depends on the level of crosslinking: k_s=1.2×10^-3 s ^-1 in the non-crosslinked CLC polymer and k_s=6.1×10^-4 s ^-1 in the very weakly crosslinked system, until a very low k_s=5×10^-5 s ^-1 in the CLC elastomer. We observe that in all materials the selective reflection peak red-shifts with time. Since the contracted helix has to endure the Frank elasticity energy penalty for the excess twisting, the system should have the tendency to recovering the initial pitch defined by the natural twisting power of the chiral material. To achieve this the system can gradually unwind the helix. This is not the case when the sample is fully crosslinked since its number of helical turns is frozen into the network. Therefore, it is reasonable that the crosslinked CLC polymer shows much slower relaxation than the non-crosslinked CLC system. Also, in the stretched elastomer the equilibrium peak position at t→∞, given by the constant B=0.74, is close to the instantaneous band shift, while in the polymer the helix does return to its pre-deformation value (Δλ_c→0) after a long time. It is important to note that the crosslinking did not affect the fast relaxation process shown in the beginning: this suggests that some common factors, such as the relaxation of the supporting silicon elastomer, dominate this fast relaxation.

 

Fig. 8. Relaxation of the contracted helix under uniaxial strain after the samples of non-crosslinked and weakly crosslinked CLC polymer were instantaneously stretched by 70%. The plot shows the normalized shift in the R* selective reflection peak position; the data is fitted by a double exponential equation. The dashed line suggests the level that would be exhibited in the fully crosslinked CLC network.

Download Full Size | PDF

Stimulated laser emission was verified by using the three-layer structure (an isotropic elastomer containing dye sandwiched between two cholesteric elastomers, acting as distributed feedback mirrors forming the cavity between them) that is extensively described in the experimental section. Rhodamine B dye is dispersed in the isotropic medium which constitutes the intermediate layer: this dye exhibits a maximum of absorption around 560 nm and a maximum of emission around 605 nm. In our all-elastomer device, we do not need to be precise about matching the bandgap position and the dye emission range. When the sample is mounted on the stretching device, we first apply uniaxial deformation to achieve: (i) the maximum of Rhodamine emission roughly in the middle of the cholesteric bandgap, and (ii) enhancing the cholesteric mirrors efficiency by “activating” the opposite-handed reflection. When pumped with a circularly polarized pulsed laser, a stimulated emission is observed, as shown in Fig. 9. The laser emission shown in the plot was obtained when the sample stretching was around 60%.

 

Fig. 9. Single mode laser emission from the CLC-MS-CLC three layer system, stretched by about 60%. Superposed is the transmittance spectrum of the CLC layer stretched by the same amount between silicone supports. The laser emission intensity is expressed in arbitrary units.

Download Full Size | PDF

In order to understand origin of the stimulated lasing we overlapped the emission with the transmittance of the cholesteric elastomer at the same elongation (both spectra are shown in the same Fig. 9): clearly, the results show the emission line in the middle of the selective reflection band where propagation of the light is restricted. This is expected since the flexible cholesteric liquid crystal layers in this configuration act as effective mirrors and the resonance in the cavity is most efficient near the middle of the reflection band [30].

Tuning of the laser peak with stretching was also achieved with this sample configuration, as illustrated in Fig. 10. It is important to point out that no laser emission could be observed when the sample was unstretched, because the Rhodamine fluorescent emission and the selective reflection gap of the cholesteric elastomer do not overlap. The lasing was observed only when the sample elongation was around 45% and the achieved tuning range was 36 nm, from 592 to 628 nm. After these measurements our samples were highly damaged and no more lasing could be observed. In the region between 585 and 650 nm the Rhodamine B dye emission spectrum is more than 50% of its maximum value, therefore we can say that the tunability has covered around 50% of the available range. The range of tunability could be increased by making the system more resistent and stable. In Fig. 10 the shift of the selective reflection band center is also shown. It appears that there is a direct proportionality between the bandgap and the stimulated emission wavelength, as expected.

Finkelmann et al. achieved a tunability range of 50 nm with the single cholesteric elastomer and the laser dye DCM [22]; more recently, Schmidke et al. measured a range of about 100 nm in an analogous system [23]. Their work also confirm the direct proportionality between the transmission band gap center and the lasing emission wavelength. To our knowledge, this is the first quantitative report on a three-layer configuration laser tunability.

We could only crudely estimate the threshold energy for lasing emission, obtaining the energy density of 7 mJ/cm² as the minimum pumping intensity to achieve emission while moving the pumping spot across the sample. This is much lower than 191 mJ/cm², the threshold energy in dye-doped single cholesteric elastomer of Schmidtke et al. [31]. However, our estimated threshold is similar to the report of Song et al. [32], who compared threshold energy in a dye-doped single cholesteric medium and in a three-layer structure where two cholesteric LC polymer films were used to sandwich a dye-doped nematic active medium. They also reported a low threshold: 3 mJ/cm² in the three-layer structure, comparing with the dye-doped single cholesteric medium of 8 mJ/cm². More recently, Takanishi et al. achieved a very low threshold with a three-layer configuration [33]. These very low threshold were obtained in samples which did not allow mechanical tunability. It is also worth noticing that the pumping laser emission (532 nm) was not exactly at the maximum of dye absorption (560 nm): this reduces the dye emission and therefore increases the effective lasing threshold.

 

Fig. 10. Wavelength of the primary emission line is plotted against sample elongation (*), and compared with the peak wavelength λ_c of the cholesteric selective reflection gap (△). Horizontal dotted lines show the range of emission spectrum of Rhodamine B (RhB) where the emission intensity is more than 50% of its maximum value (585–650 nm).

Download Full Size | PDF

It appears that cholesteric liquid crystal elastomers, confined between optically transparent silicone layers for mechanical support, can be used for tunable laser applications. Spectroscopic measurements with circularly polarized light on crosslinked samples show a behavior in qualitative agreement with the theoretical model. Surprisingly, the non-crosslinked sample shows an evident movement of the bandgap, indicating that there must be constrains (entanglement) that induce a change in the helix pitch. Studies of relaxation dynamics suggest that this is indeed a temporary, although very long-lived effect. Importantly, we have shown that two cholesteric elastomers films act as deformable and mechanically tunable “mirrors”, allowing an isotropic dye-doped layer in between to generate tunable emission in the resulting optical cavity.