3.1. Pumped-Position Controlled Random Lasing Emissions
The excited lasing patterns of the capillary tube laser 1 are shown in
Figure 2. The images of the lasing patterns were displayed on a curved screen and captured by a camera. A tiny hole was drilled in the curved screen for the pulse laser to pass through the screen and pump the capillary tube laser, which was put in the center of the curvature of the curved screen, as shown in
Figure 2. The polarization of the pulse laser was set to be in parallel with the axial direction of the capillary tube for the optimal pumped condition. When the sample was set exactly on the focus of the lens, the pattern of the excited lasing emission shows two yellow spots along the normal direction on the screen, as shown
Figure 2a. When the focused pulse beam was adjusted to deviate from the center of the DDNLC-infiltrated capillary tube, instead of the yellow spots, the pattern of the excited lasing emission showed an orange stripe on the curved screen, as shown in
Figure 2b.
Figure 3a shows the emission spectra of the capillary tube laser pumped with various pulse energies on-center. Only weak fluorescence can be excited if the pumped energy of the incident pulse is low. When the pumped energy of the incident pulse gradually increased, the emission spectra exhibited multiple peaks around 580 nm. The multi-peak emissions shown in
Figure 3a were confirmed as random lasing emissions since we got a similar result when the DDNLC was placed in a wedge cell, which does not support the existence of a Fabry–Perot resonant mode, and for which the fluorescence can be enhanced only via the closed loop formed by the random scattering [
18]. To further verify that no Fabry–Perot effect occurred in the random lasing generation, the Fourier power transforms of the lasing spectra (the red curve) of
Figure 3a were calculated and are presented in the inset. It is well-known that Fourier analyses performed for random laser spectra are represented on the 1/λ scale, so the frequencies of the obtained Fourier spectra are on the micrometer scale [
24]. The nonperiodic and broad spectral components shown in the inset of
Figure 3a indicate that the random lasing generation did not result from the Fabry–Perot effect. The generation of random lasing emission in the DDNLC micro tube is attributed to the recurrent multi-scattering and weak localization of fluorescence photons when the DDNLC region is thick enough [
20]. In this paper, the inner diameter of the capillary tube was 200 μm, which is sufficient to show spatial fluctuation in the orientational order and dielectric tensor of LCs [
18,
19]. The central wavelength of the random lasing emission of the DDNLC micro tube was around 580 nm because the laser dye P597 has the lowest energy threshold at the wavelength around 580 nm when it is doped with nematic liquid crystals [
25]. The variations in the peak intensity and full-width at half-maximum (FWHM) of the random lasing emission profile on the incident pumped energy are summarized in
Figure 3b. It can be clearly observed that when the incident pumped energy exceeded 1 µJ/pulse, the peak intensity of the random lasing emission clearly increased, while the FWHM of the emission spectrum decreased sharply. That is, the energy threshold of the random lasing emission was around 1 µJ/pulse.
Figure 3c shows the lasing spectra of the capillary tube laser when the pumped pulse was focused on DDNLC near the glass sheath of the capillary tube with various pulse energies. Similarly, only weak fluorescence could be excited if the pumped energy of the incident pulse was low. When the pumped energy of the incident pulse beam gradually increased, the emission spectra exhibited multi-peaks around 600 nm and the peak intensity increased abruptly. One possible reason for the wavelength shifting from 580 nm to 600 nm is due to boundary constraints at the capillary wall [
22]. The boundary constraints have influence on the order parameters of the dyes and the liquid crystal molecules, the total gain profile, and thus, the lasing wavelength [
26]. The variations in the peak intensity and FWHM of the lasing emission profile on the incident pumped energy are summarized in
Figure 3d.
Figure 3d indicates that when the incident pumped energy exceeded 0.8 µJ/pulse, the peak intensity of the lasing emission clearly increased, while the FWHM of the emission spectrum decreased sharply. Therefore, we can define the energy threshold of the lasing emission as 0.8 µJ/pulse.
As described in
Figure 2 and
Figure 3, the wavelength of the lasing emission depends on the pumped light being focused on or deviating from the center of the DDNLC region. Here, we put the DDNLC-infiltrated capillary tube on a translation stage to further investigate the position-dependent lasing emission. Referring to
Figure 1, the DDNLC-infiltrated capillary tube was placed on the xy-plane and aligned along the x-axis, and the incident pulse was aligned to propagate along the z-axis. The pumped location, and thus the lasing emission, can be adjusted by moving the capillary tube along the z-axis or y-axis.
First, we discuss the case of adjusting the relative distance between the capillary tube laser and the lens by moving the capillary tube along the z-axis and defining it as the shifted distance. As shown in
Figure 4a, if the shifted distance between the center of the tube and the lens was equal to the focal length, the random lasing emission with a central wavelength of 580 nm (green curve) could be excited. If the focused spot was moved away from the center of the capillary tube by 1 mm, the intensity of the random lasing emission decreased, as shown by the blue and yellow curves. If the shifted distance between the focused spot and the center of the capillary was 2 mm, a weak emission with a central wavelength of 600 nm could be excited, as shown by the orange and cyan curves. The stronger lasing emission (red and violet curves) could be measured if the center of the capillary deviated from the focused spot by 3 mm. It should be noted that the pumped pulse propagated through the capillary, which has spherical boundaries and can be regarded as a lens. Therefore, the shifted distance between the lens and the capillary tube was not equal to the variation of the focused position. After observing the variation of the lasing emission by adjusting the pumped region along the z-axis, the variation of the lasing emission by changing the pumped location along the y-axis was also of interest. The experimental results on the variation of lasing emission with adjustment of the pumped region along the y-axis is shown in
Figure 4b. The central wavelength of the random lasing emission is located at 580 nm (green curve), which occurred when the focused spot was in the center of the capillary tube. If the focused spot was moved up or down from the center of the capillary cylinder by 10 μm, weak lasing emissions with central wavelengths of 580 and 600 nm could be observed simultaneously (blue and yellow curves). Since the spherical boundaries of the capillary can be regarded as a lens with spherical aberration, which indicates that the light passing through along the central axis and off-axis will be focused on different spots, the random lasing emission at the excited location along with the z-axis (along the central axis) was a little bit different to that of y-axis (off-axis). By moving the focused spot further away from the center of the capillary cylinder, the intensities of the lasing emissions with central wavelengths of 580 and 600 nm decreased and increased, respectively.
3.2. Complete Optical Control of Lasing Emissions
When the DDNLC1 was replaced by DDNLC2, which included 8
wt% of azo-dye, the properties of the micro tube laser could be further controlled via complete optical control. The random lasing emission in the DDNLC2-infiltrated micro tube could be controlled, making it decrease with the irradiation of the UV light (365 nm) at various irradiation times t
UV = 0, 10, 20, 30, and 40 s, as shown in
Figure 5a. Here, the irradiated intensity of the UV light was fixed to 59 mW/cm
2 and the pumped energy was fixed to 9 μJ/pulse. Following the UV irradiation, one green light (532 nm) with a fixed intensity of 267.4 mW/cm
2 was employed to irradiate the tube with irradiating times of t
G = 0, 5, 15, and 20 s, as shown in
Figure 5b. The intensity of the random lasing emission could be controlled and made to return to its original value by irradiating the green light for 20 s. The variation of the random lasing intensity was attributable to the UV- and green light-induced isothermal phase transitions between the nematic and isotropic phase of the DDNLCs, via isomerization of the azo-dye. The azo-dye, 4MAB, stably exists in trans-form in darkness. The trans-4MAB dyes were roughly aligned with the molecules of the LC via the guest–host effect in the DDNLC. The trans-4MAB dyes absorbed UV light (365 nm) and transformed to a bent cis-form, resulting in the disorder of the LC host. By increasing the time of UV irradiation, t
UV, the concentration of the cis-form increased such that the phase of the LCs gradually changed from nematic to isotropic isothermally [
27]. This process caused the spatial nonuniformity of the order [δS = S(r + δr)–S(r)], and thus of the dielectric tensor [δε = ε(r + δr) − ε(r)] of the LCs, to gradually decrease from a nonzero (δS ≠ 0 and δε ≠ 0 at nematic phase) to a zero (δS = 0 and δε = 0 at isotropic phase) value, where δr, δS and δε denote the differential displacement, differential order and differential dielectric tensor of the LCs between two adjacent local micro-domains of the LCs, respectively. Therefore, the local multiple micro-domain of the LCs experienced by the propagation of fluorescence photons gradually disappeared, and thus the recurrent multi-scattering of the fluorescence decreased. When the cis-4MAB dyes were exposed to green light, the bent cis-4MAB transformed back to a trans-form and the DDNLC consequently changed back to the nematic phase from the isotropic phase. With the reappearance of the nematic phase, the uniform orientations of the LCs, and thus the recurrent multi-scattering of the fluorescence, resulted in the recovery of random lasing emissions [
19].
The UV-induced isothermal phase transition of the DDNLC-infiltrated micro tube was observed with the aid of a transmitted polarizing optical microscope (T-POM) with crossed polarizers. As shown in
Figure 6, the angle between the transmission axis of the polarizer and the axis of the capillary cylinder was 45°. By increasing the duration of the UV light irradiation, t
UV, from 0 to 40 s, the transmission decreased from the bright to the dark state, as shown in
Figure 6a. On the contrary, by increasing the duration of the green light irradiation, t
G, from 0 to 20 s, the transmission recovered back to the bright state, as shown in
Figure 6b. These results coincide with the mechanisms that cause the reversible variation of the random lasing intensity—that is, the UV- and green beam-irradiation, which induce isothermal nematic to isotropic and isotropic to nematic phase transitions of LCs via trans–cis and cis–trans isomerizations of azo-dye, respectively. During the photo-induced phase transition, the variation in temperature of the DDNLC-infiltrated micro tube was monitored with a thermal imager (Fluke, Ti10). The temperature of the micro tube only increased from 23.8 °C to 27.9 °C after irradiation with UV for 50 s, as shown in
Figure 6c,d. The final temperature, 27.9 °C, was much lower than the temperature of the clearing point of the DDNLC2, which was about 50 °C. This result indicates that no thermal-induced phase transition occurred during the UV irradiation process. As a result, we can ascertain that the complete optical control of the lasing emission was not caused by thermal effects but by the effects of photoisomerization.
As discussed in
Section 3.1, the wavelength of the lasing emission can be changed from 580 nm to 600 nm if the focused pulse deviates from the center of the micro tube. In this section, we adjusted the position of the focused spot (off-center pumped) to generate a lasing emission with a central wavelength of 600 nm. When the micro tube was irradiated by the UV light at various irradiation times t
UV = 0, 10, 20, 30, 40, and 45 s, the variations of the emission spectra were as shown in
Figure 7a. At t
UV = 0, 10, and 20 s, the central wavelength of the lasing emission was located around 600 nm and its intensity gradually decreased. At t
UV = 30 s, two groups of lasing emissions with main wavelengths located at 600 nm and 620 nm could be observed on the spectrum. At t
UV = 40 and 45 s, the central wavelength of the lasing emission was at 620 nm and the emission intensity gradually increased with t
UV. Following the UV irradiation, the micro tube was irradiated by green light with the irradiation time t
G = 0, 5, 15, 20, and 25 s to trigger the cis to trans isomerization of the azo-dye. As shown in
Figure 7b, the wavelength and intensity of the lasing emission could be controlled and made to return to its original value by irradiation the green light. It is interesting to note that the random lasing emission shown in
Figure 5 could be switched off after 40 s of UV light irradiation, but that the off-center pumped lasing emission exhibited not only the decrease in intensity, but also the change of the wavelength after 40 s of UV light irradiation. The difference in the irradiation duration of the control light between
Figure 5 and
Figure 7 was mainly attributable to the boundary constraints at the capillary wall, since the lasing emissions shown in
Figure 5 and
Figure 7 were from the buck region and the region near the glass sheath of the capillary tube, respectively.
The phase transition was the main reason for the wavelength change and higher intensity of the lasing emissions. In a laser, the total gain can be expressed as
G(
λ) =
g(
λ)
ℓ −
σ(λ)L, where
ℓ is the illumination length,
L is the internal circumference of the resonant path,
g(
λ) is the gain of the laser dye for unit length, and
σ(
λ) is the loss for unit length in the system [
23]. As the total gain at 600 nm was higher than that at 620 nm when the DDNLC was in the nematic phase, the central wavelength of the lasing emission at the nematic phase was around 600 nm. On the contrary, the total gain at 620 nm exceeded that at 600 nm and thus the wavelength of the lasing emission shifted to 620 nm when the DDNLC was in the isotropic state. In addition to the shift of the lasing emission, the intensity of the lasing emission was higher at the isotropic phase compared to that at the nematic phase. Since the absorption of the laser dye and scattering from the thermal fluctuations can be significantly reduced when the DDNLC is in isotropic phase [
28], the emission intensity at the isotropic state was higher than the nematic state.