1. Introduction

Random laser has received considerable research attention since it was proposed by Letokhov in the late 1960s [1]. Unlike traditional laser, the generation of random laser is free of a resonance cavity with high-quality reflection mirrors. The latent cavities for random laser are formed by the feedback of photons due to multiple scattering in disordered gain-scattering media [2–5]. When the scattered photons are amplified efficiently in the latent cavities and make the gain larger than the loss in the system, lasing will occur. The combination between scattering and amplification triggers the random laser. Random laser possesses distinct feature owing to its unique excitation mechanism involving low spatial coherence and simplicity. Hence, random laser has many potential applications, including nanotechnology, full field imaging and projection [6]. The amplified-scattering mechanism has been realized in numerous materials, such as semiconductor powder and those of dye-doped liquid crystals, biomaterials or polymer [4,7–10].

Metal nanoparticles (NPS), due to their unique property of localized surface plasmon resonance (LSPR) and enhanced scattering characteristic, have been widely doped in the random gain media to adjust coherent and incoherent random lasing [11–19]. On the one hand, the LSPR will enhance the localized electromagnetic field around the metal NPS. On the other hand, compared to dielectric scatters, the metal NPS have larger scattering cross section for visible light and can result in enhanced scattering strength within these random media. An incoherent random emission in dye-doped Ag NPS random gain system was firstly reported by Dice et al. [12]. Following the route, Popov et al. showed the enhanced lasing characteristic by embedding Au NPS in a thin polymer film containing dyes [13]. By introducing metal NPS with different material, shape, dimension and aggregation, the effect of metal NPS on random lasing was concretely investigated [14–16]. Recently, Ye et al. studied the variation of threshold, turn-on time and turn-off time of the random laser by varying the concentration of Ag NPS in dye-doped nematic liquid crystals [8]. An ultra-thin plasmonic random laser which consists of a free-standing polymer membrane embedded with Ag NPS was realized by Zhai T.R et al. [17], and they also fabricated a mechanically-tunable random laser based on a waveguide-plasmonic scheme [18]. Hu Z.J et al. reported the random fiber laser based on the LSPR of Au NPS in liquid core optical fiber [19].

Recently, we reported random lasing in a disordered system that made up of manganese chloride and dye-doped polymer dispersed liquid crystals. Low-threshold and high-intensity random laser observed in the system is based on the energy transfer between dye molecules and manganese chloride molecules [20]. As we know, due to the random walk of the photons, many different latent cavities are formed in the random gain systems. So the emission of the random systems is not in a specific direction, it can be observed in complete solid angle. However, reports about this perfect emission are few. In this work, we investigated the random lasing from dye Pyrromethene-597 (PM597) doped supersaturated copper sulfate (CuSO₄) capillaries samples through varying the concentration of CuSO₄ or adding Au NPS as extra light scatters. The results show that random laser can be observed in all detected direction around the PM597-doped CuSO₄ sample, meanwhile, the threshold and energy of random laser can be tuned neatly by varying the CuSO₄ concentration. When Au NPS is doped in the sample, the threshold of random laser is lowered and the number of lasing modes is reduced. Our demonstrated approach not only indicates CuSO₄ is a favorable scatters in random gain systems but also pave a new avenue towards designing low-cost optical devices based on random laser.

2. Experiments

Pyrromethene-597 (PM597) dispersed in ethanol was chosen as optical gain material in this work. Solid $CuS{O}_4•5H_{2}O$ (7.354 g) was dispersed into deionized water (10 mL) to form CuSO₄ supersaturated solution. We prepared four uniform mixtures of PM597-doped CuSO₄, which contained 0.3 mg PM597, 0.2 mL ethanol and different volumes of CuSO₄ supersaturated solution (0.1, 0.2, 0.25 and 0.3 mL). For comparison, we added 0.1 mL ethanol solution of Au NPS into the above PM597-doped CuSO₄ (with 0.2 mL CuSO₄), where the number density of the Au NPS is $1.025\times {10}^{11}/mL$. Figure 1(b) illustrates the morphology of spherical Au NPS we used, revealing the average diameter is 50 nm. All the mixtures were shaken for 2 hours by ultrasonic and poured into glass capillaries, as illustrated in Fig. 1(c), which is of 50 mm in length and 0.5 mm in diameter. The ingredient of all the samples is shown in Table 1.

 

Fig. 1 (a) Experiment setup for measuring random lasing. (b) TEM images of Au NPS with average diameter 50 nm. (c) Schematic diagram of the ingredients in sample. (d) The normalized absorption intensity of PM597, Au NPs, CuSO₄ and fluorescence intensity of PM597.

Download Full Size | PDF

Table 1. The ingredient of all the samples.

View Table

The experimental scheme for emission measurement is shown in Fig. 1(a). The pump beam from a second-harmonic of a Q-switched Nd: YAG laser (532 nm wavelength, 10 HZ repetition rate and 8 ns pulse duration) was divided into two sub-beams by a polarizing beam splitter. The reflected sub-beam emission is collected as a reference beam by energy meter while the transmitted sub-beam, after propagating a half-wave plate, is focused on the capillary samples by a lens and a slit. Emission from the capillary sample’s surface was collected with an optical fiber bundled into a spectrometer to resolve the spectra.

3. Results and discussion

Figure 2(a) plots the evolution of emission spectra with pump energy for PM597 system. Broad fluorescence spectra with a full width at half maximum (FWHM) of about 20 nm can be observed under different pump energies. The fluorescence intensity is enhanced with the increment of the pump energy. The evolution of emission spectra against pump energy for PM597-doped CuSO₄ (0.1 mL) is shown in Fig. 2(b). At low pump energy, there is only a broad spontaneous emission spectrum centered at 604.8 nm with the FWHM of about 7.5 nm. A single sharp peak centered at 607.7 nm with the FWHM of about 0.23 nm over the broad background spectrum appears when the pump energy increases to 9.4$\mu J$. As the pump energy increases further to 16.4$\mu J$, multiple sharp peaks emerge with the FWHM much less than 0.20 nm. Moreover, the number of the sharp peaks increases with the pump energy increasing, indicating the occurrence of coherent random lasing. This is closely related to CuSO₄ micro-crystal molecules that not fully dispersed in the mixture sample, which causes multiple scattering of photons in the sample. Due to the multiple scattering and wave interference, the photons may be trapped inside the gain medium and form different optical feedback loops of photons. The photons are amplified in the optical feedback loop and exceed the loss, implying the occurrence of a random laser. Because of the random distribution of scattering particles inside the PM597-CuSO₄ mixture, the optical feedback loops of photons in it will be random, resulting in multiple lasing modes. When the pump energy increases, gain greater than loss appears in more optical feedback loop, which results in more discrete peaks in the emission spectrum. In addition, as shown in Fig. 1(d), the emission spectrum of PM597 overlaps with the absorption spectrum of CuSO₄, which indicates that the random lasing phenomenon in the PM597 doped CuSO₄ (0.1 mL) is also associated with the energy transfer between CuSO₄ molecules and PM597 molecules. The photons emitted by dye PM597 is subsequently absorbed by CuSO₄ molecules. Besides, the dye molecules may directly transfer its excitation energy to a ground-state of the CuSO₄ molecules [21]. The pump-dependence of the integrated emission intensities for PM597, as depicted in Fig. 2(c), do not exhibit threshold behavior, implying spontaneous emission of PM597 molecules. However, the pump-dependence of the integrated emission intensity in PM597-doped CuSO₄ (0.1 mL) shows a clear threshold behavior, which is a typical characteristics involved in random lasing with coherent feedback. Figure 2(d) plots single-shot emission spectra of PM597 doped CuSO₄ (0.1 mL) system at the pump energy of 16.65$\mu J$, where the laser spikes as well as emission intensities exhibit chaotic behavior, indicating pump-dependent competitions among different laser multimodes.

 

Fig. 2 The evolution of emission spectra of (a) PM597 and (b) PM597-doped CuSO₄ (0.1 mL) at different pump energies. (c) Dependence of emission intensity on the pump energies for PM597 and PM597-doped CuSO₄ (0.1 mL). (d) Single shot lasing spectra of PM597-doped CuSO₄ (0.1 mL) at the pump energy of 16.65 $\mu J$.

Download Full Size | PDF

Figure 3(a) depicts emission spectra versus different detection angles $\theta$ (see Fig. 1(c)) for PM597-doped CuSO₄ (0.25 mL) system. It is found that lasing modes are observed at the complete solid angle, however, lasing modes are different at different angles. Since the different optical feedback loops could have different output directions, and may have stronger emission into some directions. The peak intensity of lasing signal as a function of detection angle is presented in Fig. 3(b), it is noted that the intensity of lasing signal is highly dependent on the detection angle and reaches the maximum at the angle of $\theta ={90}^0$. This is reasonable since the angle of $\theta ={90}^0$ is the incident direction of the pump light, corresponding to the most efficient illumination of pump light. Moreover, in this direction light may experience the longest optical path of gain and can be absorbed much more fully than in other directions, reaching the strongest output intensity of random lasing.

 

Fig. 3 (a) Emission spectra of PM597-doped CuSO₄ (0.25 mL) recorded at different detection angles. (b) Dependence of peak emission intensity on the detection angles for PM597-doped CuSO₄ (0.25 mL).

Download Full Size | PDF

Single-shot emission spectra of PM597-doped CuSO₄ with the concentration of CuSO₄ 0.1 mL, 0.2 mL and 0.3 mL are illustrated in the inset of Fig. 4(b), where the samples are pumped with the same pump energy of 12.00$\mu J$. Red shift of the main peak wavelength is observed with increasing the concentration of CuSO₄ from 0.1 mL to 0.3 mL. It is responsible for the thicker density of CuSO₄ that supports, for relatively redder photons, more efficient energy transfer between the dye PM597 molecules and those of CuSO₄. In particular, we find that the spectrum of PM597 doped CuSO₄ (0.3 mL) appears two main peaks centered at 611.79 and 627.28 nm respectively, which seems to be a manifestation of energy level splitting in the coupled-resonance transition of dye and CuSO₄ molecules.

 

Fig. 4 (a) Emission intensity as a function of the pump energy for PM597-doped CuSO₄ (0.1, 0.2 and 0.3 mL). (b) Ensemble-averaged Power Fourier Transform (PFT) analysis for single-shot emission spectra (sum = 50) of PM597-doped CuSO₄ (0.1, 0.2 and 0.3 mL) at the pump energy of 12.00$\mu J$. Inset: Two individual single-shot emission spectra recorded for the PM597-doped CuSO₄ (0.1, 0.2 and 0.3 mL) at the pump energy of 12.00$\mu J$.

Download Full Size | PDF

Figure 4(a) plots the dependence of integrated emission intensities upon the pump energies for the three PM597-doped CuSO₄ mixture samples. All the samples manifest threshold behavior and have thresholds of 8.9, 7.2 and 9.25$\mu J$, corresponding to concentrations of CuSO₄ 0.1, 0.2 and 0.3 mL, respectively. The sample with CuSO₄ 0.2 mL has the lowest threshold, which is attributed to the highest cavity quality factor of the random laser in it. The quality factor $Q$ is a parameter defined commonly to exhibit the peculiarity of laser cavity, the higher value it is, the less loss the resonant cavity will be. For mirrorless resonant cavity of random laser, the quality factor is given by [22,23]

To further verify the above judgement, we calculated the ensemble-averaged Power Fourier Transform (PFT) analysis (see Fig. 4(b)) for a sum of single-shot emission spectra of the three PM597-doped CuSO₄ (0.1, 0.2 and 0.3 mL) mixture samples, respectively. All the single-shot emission spectra are collected at various sample positions upon the same pump energy of 12.00 $\mu J$. Positions of the peaks in PFT spectra are expressed as follows [24,25],

Figure 5 demonstrates the effect of the pump energy on random lasing spectra and pump–emission intensity dependence for the samples PM597-doped CuSO₄ (0.2 mL) without and with Au NPS. For PM597-doped CuSO₄ (0.2 mL), the emission spectra of random laser at different pump energy is shown in Fig. 5(a), where a random lasing main peak centered at 610.2 nm with the FWHM of about 0.21 nm emerges at the pump energy of 7.65$\mu J$. As a comparison, random lasing action of PM597-doped CuSO₄ (0.2 mL) with Au NPS is depicted in Fig. 5(b), which exhibits a sharp peak located at 602.1 nm with the FWHM of about 0.23 nm at the pump energy of 5.96$\mu J$. Figure 5(c) plots the evolution of the integrated emission intensities of the output random lasers with pump energies, corresponding to the PM597-doped CuSO₄ (0.2 mL) without and with Au NPS, respectively. It shows in the both systems that the emission intensities increase gradually with the increase of the pump energies, where an abrupt increase of the emission intensities occurs, suggesting the threshold behavior of the both systems. In addition, thresholds of the former and the latter systems are 7.20 and 5.26$\mu J$, respectively, indicating that the threshold is reduced as doping Au NPS in the PM597-doped CuSO₄ (0.2 mL) sample. This is owing to the localized surface plasmon resonance (LSPR) and the larger scattering cross section caused by Au NPS. The LSPR enhances the localized electromagnetic field around the Au NPS which facilitates the simulated emission of laser dye, while the larger scattering cross section provides stronger multiple scattering of light which result in longer resonance cavity length of random laser and reduces the pumping threshold.

 

Fig. 5 The evolution of emission spectra of PM597-doped CuSO₄ (0.2 mL) (a) without and (b) with Au NPS at different pump energies. (c) Dependence of emission intensity on the pump energies for PM597-doped CuSO₄ (0.2 mL) without and with Au NPS. (d) Ensemble-averaged PFT analysis for single-shot emission spectra (sum = 50) of PM597-doped CuSO₄ (0.2 mL) without and with Au NPS at the pump energy of 12.75$\mu J$. Inset: Two individual single-shot emission spectra recorded for the PM597- doped CuSO₄ (0.2 mL) without and with Au NPS at the pump energy of 12.75$\mu J$.

Download Full Size | PDF

Similar to Fig. 4(b), ensemble-averaged PFT curves for a sum of (sum = 50) single-shot emission spectra of PM597-doped CuSO₄ (0.2 mL) without and with Au NPS at the same pump energy of 12.75$\mu J$are plotted in Fig. 5(d). The inset of Fig. 5(d) depicts two individual single-shot emission spectra of PM597-doped CuSO₄ (0.2 mL) without and with Au NPS at the same pump energy of 12.75$\mu J$. The single-shot emission spectral of the sample with Au NPS appears a clear blue-shift with respect to that of the sample without Au NPS, which is due to the variance of fluorescence quenching efficiency caused by the LSPR of Au NPS. In addition, it is worth noting that the number of lasing modes in the sample with Au NPS is less than that in the sample without Au NPS. This can also be observed in the inset of Fig. 5(d), in which the PFT curve of the sample with Au NPS has fewer number of the Fourier transform harmonics. When Au NPS is doped in the sample, a longer scattering mean free path results in less number of resonant cavities where the lasing threshold can be reached [27]. According to Fig. 5(d), we obtain the oscillation cavity lengths of the samples without (8.60$\mu m$) and with Au NPS (26.23$\mu m$). Evidently, the resonant cavity length of the sample is increased by three times when Au NPS is doped in it, the longer cavity length means higher quality factor of the cavity length or lower threshold of random laser. Besides, the partial overlap between absorption spectrum of Au NPS (d = 50 nm) and fluorescence emission spectrum of PM597 dyes, as plotted in Fig. 1(d), implies that the photon emitted by PM597 dye molecule can be absorbed by Au NPS. This mechanism can also enhance the LSPR effect of Au NPS and extends the lifetime of laser photons in the system.

Contour maps of shot-to-shot lasing spectra at different pump pulses for PM597-doped CuSO₄ (0.2 mL) without and with Au NPS are depicted in Fig. 6(a) and 6(b) while keeping pump position and pump energy (13.10$\mu J$) fixed. It is seen that the lasing in the former system (Fig. 6(a)) shows unstable frequencies but the lasing frequencies in the latter system (Fig. 6(b)) are more stable, moreover, the latter system shows higher intensity. The variable lasing frequencies in the former system is correlated to the dynamic competition of multiple laser modes. However, for the systems with Au NPS, the enhanced LSPR can increase the coherence of optical field, inducing steadier lasing modes and fewer laser spikes (see Fig. 5(d)).

 

Fig. 6 Contour maps of shot-to-shot lasing spectra at different pump pulses for PM597-doped CuSO₄ (0.2 mL) (a) without and (b) with Au NPS at the pump energy of 13.10$\mu J$.

Download Full Size | PDF

4. Conclusion

To summarize, we studied the random lasing from PM597-doped CuSO₄ glass capillaries. Results show that the threshold, lasing modes, peak intensity and emission wavelength can be tuned by changing the concentration of CuSO₄ in the mixture materials. Diverse emission spectra are observed in different detection directions around the surface of the sample. This is responsible for the multiple scattering of CuSO₄ micro-crystal dispersed in the mixture materials. In addition, the energy transfer between dyes molecules and CuSO₄ also plays an important role in the tuning properties of the random laser. The comparison of cavity lengths in the three systems confirms the effect of CuSO₄ on the random lasing. Besides, random laser with lower threshold, higher intensity and fewer modes was found when Au NPS is doped into the PM597-doped CuSO₄ sample, which is due to LSPR and a larger scattering cross section provided by Au NPS. The PFT-analysis on laser cavity lengths and the contour maps on time-frequency domain of shot-to-shot lasing spectra for both systems reveal the positive influence of Au NPS on improving random laser performance. These results not only enrich the realization schemes of general random laser but also provide insights into rules of random lasing occurrence in some complex gain media with energy transfer between their ingredients.