We report random laser action in a system where optical amplification is provided by colloidal quantum dots (CQDs). This system is obtained by depositing from solution CdSe/ZnS core-shell CQDs into rough micron-scale grooves fabricated on the surface of a glass substrate. The combination of CQD random packing and of disordered structures in the glass groove enables gain and multiple scattering. Upon optical excitation, random laser action is triggered in the system above a 25-mJ/cm2 threshold. Single-shot spectra were recorded to study the emission spectral characteristics and the results show the stability of the laser mode positions and the dominance of the modes close to the material gain maximum.
© 2011 OSA
Recently, “random” lasers have attracted great interest within the scientific community. Their concept is based upon light generation and stimulated emission within a multiple-scattering medium with optical gain . Consequently, there is no conventional optical resonator in the laser system, which makes for flexible and, in the long run, potentially low-cost device fabrication. These coherent light sources resemble conventional lasers in many ways, such as spectrally narrow emission, threshold behavior, relaxation oscillations and photon statistics [2,3]. However they have unique characteristics that are rich for fundamental studies and which could one day open up interesting applications in areas including new generation display devices, environment lighting, sensors, optical coding of objects, random number generators and medical diagnostics [1,4]. To date, phenomena of random laser action have been demonstrated in non-ordered systems formed by various materials such as rhodamine 640 with TiO2 particles , ZnO powder , polymer composites , rhodamine 6G doped in PMMA film incorporated with Ag nanoparticles  and dye-doped POSS solutions . In this work, we report what is to our knowledge a novel random laser material system based on colloidal quantum dots (CQDs).
Optical gain has been demonstrated in CQDs, and owing to their size-dependent emission wavelength over a broad range (i.e. color tunability) as well as insensitive radiative transition under various temperatures (i.e. temperature stability), they offer great potential for fabricating new types of laser devices [10–15]. Because random laser action relies on the combined effects of multiple scattering and optical amplification, CQDs certainly appear to be an attractive material system for such applications. They can play the role of gain elements while scattering is inherently obtained by the disorder in their spatial positioning and size (although size disorder could be negligible compared to spatial disorder when using monodisperse CQDs). On the processing side, CQDs, like organic chromophores, can be deposited from solution on a wide variety of substrates and can also be blended with other materials and particles, an interesting feature for random laser experiments. Furthermore, they are inorganic in nature and are therefore quite robust, making them advantageous compared to organic gain materials. However, to the best of our knowledge, there is only one report on random lasing in colloidal CdS/CdSe/CdS quantum wells under cryogenic conditions , and no dedicated study on random lasers using CQDs at room-temperature has yet been made.
The novel random laser system that we report here consists of micron-scale grooves with rough internal surfaces directly written on a glass substrate onto which CQDs are deposited from solution. The groove roughness serves two purposes: on one hand it helps trap CQDs and increases their density in order to enable stimulated emission and on the other hand it enhances light-scattering. Preparation of the system using red-emitting core-shell CdSe/ZnS CQDs is explained in section 2, while section 3 describes the optical pumping set-up. Random laser action at room-temperature in this system is demonstrated and studied in section 4.
2. Sample preparation
The sample was prepared as follows. Micron-scale grooves (width: 40-120 μm and height: 30-60 μm) were written into a 500-μm-thick glass substrate by manually pressing and translating a diamond tip on the surface, as shown in Figs. 1(a) and 1(b). Several parallel trenches were fabricated in this way. Red-emitting core-shell CdSe/ZnS CQDs (Evident Technology) having a 5-nm mean diameter were then deposited from toluene solution, at a concentration of 5 mg/ml, onto the substrate by spin-coating at a low speed (300 rpm for 120 s) to allow more CQDs to be trapped. The inside surfaces of the micron-scale trenches are quite rough [Figs. 1(a) and 1(b)]. The configuration was used here to trap the CQDs inside the groove during the formation of the film  and thus to increase the CQD density above levels enabling light amplification [10,11]. Figures 1(c) and 1(d) show micrograph pictures under UV-light illumination of the final sample, i.e. after CQDs deposition. It clearly illustrates the fact that CQDs are mainly concentrated into the grooves [Fig. 1(c)], which serve as an assembly template. Finally, the trenches’ roughness is also intended to provide multiple scattering, an essential process to trigger random laser oscillation. To a certain extent the sample can be considered as a series of large-area (about 80 μm × 40 μm) scattering optical waveguides made of a CQD layer deposited inside rough glass trenches.
3. Optical pumping experiment
The set-up for the measurements is shown in Fig. 2 . For the random laser experiments, the CQDs sample was photo-pumped by a frequency-tripled Q-switched Nd:YAG laser system yielding 5-ns pump pulses at an excitation wavelength of 355 nm and a repetition rate of 10 Hz. A cylindrical lens was aligned to shape the pump beam as a stripe with a full width at half maximum (FWHM) of 0.05 × 3.0 mm2. The pump beam was incident on the sample vertically to the substrate plane and the pump fluence could be adjusted through a combination of a half-waveplate-polarizing beam splitter and a neutral density filter. The laser pump stripe was parallel to the glass grooves (inset of Fig. 2) and, given its size, only one groove was excited at a time. Both edge and normal emission from the sample could be collected and recorded with 50-µm-core optical fibres connected to a multi-channel grating-CCD spectrometer. The spectrometer had two useable detection channels with respective spectral resolutions of 2.4 nm and 0.13 nm.
4. Random laser action
Figure 3(a) illustrates the spectral evolution with the pump fluence (spectrometer resolution: 2.4 nm). The integrated peak intensity (602-608 nm) of the edge emission is plotted in Fig. 3(b) as a function of the pump fluence. It shows the typical transition of a laser transfer function with a threshold at ~25 mJ/cm2. It is seen that above threshold a peak, centered at 606 nm, appears in the emission spectrum and increases with the pump fluence. Its intensity follows a soft threshold curve rather than a sharp threshold which is due to a significant amount of spontaneous emission coupled into the laser mode . This is a well-known and understood effect and is common in micron-sized lasers. It has been studied in more detail for example in the case of ring resonators  and distributed feedback lasers . The FWHM of the emission peak drops from 30 nm below threshold to 4 nm above 29 mJ/cm2. In Fig. 3(b), it is also worth noting that the integrated intensity increases nonlinearly when the increasing pump fluence is below 5 mJ/cm2. We attribute this effect to the saturation of the 1S excitonic absorption and to biexcitonic Auger recombination . These two effects are indistinguishable on the basis of our measurements because they result in very similar PL curves, leading to a slowly increased PL evolution above 5 mJ/cm2 before reaching a soft laser threshold at 25 mJ/cm2. Normalised edge and vertical emission spectra taken at a 29-mJ/cm2 pump fluence are superposed in Fig. 3(c). Because the pump is above threshold the edge emission spectrum is quite narrow, about 4-nm FWHM as stated previously, but the vertical emission spectrum resembles that of typical CQDs photoluminescence (PL) with a FWHM of about 30 nm. This is an indication that the edge emission is dominated by stimulated emission and that there is indeed net optical gain, but PL dominates the normal emission. We might attribute this difference to the fact that both the groove and pump geometry enable stimulated emission in the preferred direction along the groove. Consequently, we only study the edge emission in the following.
A model of random lasers based on light diffusion under inclusion of optical gain demonstrates that for a given excitation fluence, there exists a critical volume of gain material above which there is stimulated emission and below which there is none [1,21]. To verify that this effect happens in our sample, the pump stripe length, and hence the excitation volume, was varied while the edge-emission spectra were recorded. This was carried out with a fixed pump fluence of 41 mJ/cm2 and results are shown in Fig. 4 . One can see that the intensity increases dramatically when the pump stripe length reaches about 3.0 mm, corresponding to the amplification length of our CQDs random system at the given pumping level .
When the emission spectrum above threshold is detected on the high-resolution channel (0.13 nm) of the spectrometer, it is observed that the 4-nm broad stimulated emission peak actually consists of a multitude of very narrow line-width peaks. Examples of such spectra are shown in Figs. 5(a) and 6(b) recorded for a pump fluence of 53 mJ/cm2. These peaks are spectrally narrow with some of them below the spectrometer resolution and their intensities are far above the noise level. This is a typical signature of random laser action when the excited region size somewhat restricts the total number of amplified modes. If we ascribe the random laser modes to these peaks, we can say that there is strong mode competition as evidenced by the fluctuating intensities of the peaks from shot-to-shot. However, for a fixed position of the pump stripe, some of these modes, centered around the stimulated emission maximum, are seen to always dominate as shown by detecting the spectrum over several emitted pulses [Inset of Fig. 6(c)].
In a random laser, it is not surprising to see strong intensity fluctuations caused by mode competition under a nanosecond photo-pumping regime . For a random system, the number of modes (N) can be estimated by N = 2πA/λ2, where A is the excited area and λ is emission wavelength [23,24]. In our case, A ~0.15 mm2 and taking 606 nm as the central wavelength, then N ~2.6 × 106. Thus mode competition is expected to be strong and only the few modes with the longest lifetime will be amplified efficiently, the other modes suffering as a consequence of a gain reduction [3,23]. In this situation, single-shot measurements can yield a different spectrum each time a pump pulse excites the gain medium. A previous study showed that semiconductor laser action along scattering paths starting at a spontaneous emission seed will bring phase or frequency fluctuations into the laser emission. As a result, the oscillation amplitudes and gain  of the individual laser modes may change. Such an explanation is also applicable for random laser action [23,26]. To study this effect and also the dominance of some of the modes in our CQDs random laser system, we recorded single-shot spectra for a given pump stripe position on the sample. The random laser single pulse spectrum evolution when pumping at 53 mJ/cm2 is shown in Fig. 6(a) with typical spectral plots in Fig. 6(b). Shot-to-shot fluctuations are seen but the mode positions are actually quite stable although the intensities for each one are not. We added up all the (normalized) 25 single-shot spectra, showing the several dominant peaks [inset of Fig. 6(c)]. Such peaks correspond to the emitted photons in these specific modes that survive the longest time in the gain region.
Because of the refractive index inhomogeneity in the random gain medium, a simplified model using a random ring-like cavity structure has been applied successfully to explain the mechanism behind phenomena of the random laser action . In any random cavity formed by long-range fluctuations of the refractive index, total internal reflections can take place from the boundaries of the region. Localized modes with close frequencies and quality factors may be generated and it has been studied that the power Fourier transformation (PFT) of the random lasing spectrum can be a very useful method to extract the random cavity diameter in the random gain medium . Corresponding (averaging) PFT of the different spectra are shown in Figs. 5(b) and 6(c), respectively. For the single-shot results, we took the individual PFT spectrum before averaging the spectra and then averaged the PFT spectra [Fig. 6(c)]. In the equivalent cavity representation frame of random lasers with coherent feedback, the periodicity can be attributed to the formation of cavity-like structures in the disordered system . In our case the average spatial periodicity length, Δd, extracted from the PFT spectra is about 150 μm. We can then deduce the average equivalent cavity diameter D = 176 μm by using the relationship  of Δd = n•D/2, where n = 1.7 is the effective refractive index of the gain medium  at the lasing wavelength.
In summary, we have demonstrated random laser action in a CQD-based system. The system was made by depositing CdSe/ZnS core-shell CQDs into microscale grooves which were fabricated on the surface of a glass substrate. Owing to the random packing of the CQDs and disordered sub-micron structures in the glass groove, multiple optical scattering was obtained and, upon optical excitation, random lasing action occurred. Single-shot spectra were recorded to study the dynamics of the emitted laser modes and the results show the stability of the mode positions and the dominance of modes close to the gain maximum. This is to our knowledge the first experimental demonstration of random laser action at room-temperature based on CdSe/ZnS core-shell CQDs as the amplifying elements and it opens the way to further work on random lasers using this material system.
This work was supported by UK EPSRC under the project of ‘HYPIX’. Y. Chen acknowledges the support from Scottish Universities Physics Alliance (SUPA).
Refere nces and links
1. D. S. Wiersma, “The physics and applications of random lasers,” Nat. Phys. 4(5), 359–367 (2008). [CrossRef]
2. S. Gottardo, R. Sapienza, P. D. Garcia, A. Blanco, D. S. Wiersma, and C. Lopez, “Resonance-driven random lasing,” Nat. Photonics 2(7), 429–432 (2008). [CrossRef]
4. A. Uchida, K. Amano, M. Inoue, K. Hirano, S. Naito, H. Someya, I. Oowada, T. Kurashige, M. Shiki, S. Yoshimori, K. Yoshimura, and P. Davis, “Fast physical random bit generation with chaotic semiconductor lasers,” Nat. Photonics 2(12), 728–732 (2008). [CrossRef]
5. N. M. Lawandy, R. M. Balachandran, A. S. L. Gomes, and E. Sauvain, “Laser action in strongly scattering media,” Nature 368(6470), 436–438 (1994). [CrossRef]
6. H. Cao, Y. G. Zhao, S. T. Ho, E. W. Seelig, Q. H. Wang, and R. P. H. Chang, “Random laser action in semiconductor powder,” Phys. Rev. Lett. 82(11), 2278–2281 (1999). [CrossRef]
7. R. C. Polson, A. Chipouline, and Z. V. Vardeny, “Random lasing in π-conjugated films and infiltrated opals,” Adv. Mater. 13(10), 760–764 (2001). [CrossRef]
8. X. Meng, K. Fujita, S. Murai, and K. Tanaka, “Coherent random lasers in weakly scattering polymer films containing silver nanoparticles,” Phys. Rev. A 79(5), 053817 (2009). [CrossRef]
9. A. Costela, I. Garcia-Moreno, L. Cerdan, V. Martin, O. Garcia, and R. Sastre, “Dye-doped POSS solutions: random nanomaterials for laser emission,” Adv. Mater. 21(41), 4163–4166 (2009). [CrossRef]
10. V. I. Klimov, A. A. Mikhailovsky, S. Xu, A. Malko, J. A. Hollingsworth, C. A. Leatherdale, H. Eisler, and M. G. Bawendi, “Optical gain and stimulated emission in nanocrystal quantum dots,” Science 290(5490), 314–317 (2000). [CrossRef] [PubMed]
11. H. Eisler, V. C. Sundar, M. G. Bawendi, M. Walsh, H. I. Smith, and V. Klimov, “Color-selective semiconductor nanocrystal laser,” Appl. Phys. Lett. 80(24), 4614 (2002). [CrossRef]
12. V. C. Sundar, H. Eisler, T. Deng, Y. Chan, E. L. Thomas, and M. G. Bawendi, “Soft-lithographically embossed, multilayered distributed-feedback nanocrystal lasers,” Adv. Mater. 16(23-24), 2137–2141 (2004). [CrossRef]
13. P. T. Snee, Y. Chan, D. G. Nocera, and M. G. Bawendi, “Whispering-gallery-mode lasing from a semiconductor nanocrystal/microsphere resonator composite,” Adv. Mater. 17(9), 1131–1136 (2005). [CrossRef]
16. J. Xu and M. Xiao, “Lasing action in colloidal CdS/CdSe/CdS quantum wells,” Appl. Phys. Lett. 87(17), 173117 (2005). [CrossRef]
17. T. Kraus, L. Malaquin, H. Schmid, W. Riess, N. D. Spencer, and H. Wolf, “Nanoparticle printing with single-particle resolution,” Nat. Nanotechnol. 2(9), 570–576 (2007). [CrossRef]
18. J. Herrnsdorf, B. Guilhabert, Y. Chen, A. Kanibolotsky, A. Mackintosh, R. Pethrick, P. Skabara, E. Gu, N. Laurand, and M. Dawson, “Flexible blue-emitting encapsulated organic semiconductor DFB laser,” Opt. Express 18(25), 25535–25545 (2010). [CrossRef] [PubMed]
19. Y. Boucher and P. Féron, “Generalized transfer function: A simple model applied to active single-mode microring resonators,” Opt. Commun. 282(19), 3940–3947 (2009). [CrossRef]
20. V. I. Klimov VI, A. A. Mikhailovsky, D. W. McBranch, C. A. Leatherdale, and M. G. Bawendi, “Quantization of multiparticle auger rates in semiconductor quantum dots,” Science 287(5455), 1011–1013 (2000). [CrossRef] [PubMed]
22. J. Fallert, R. J. B. Dietz, J. Sartor, D. Schneider, C. Klingshirn, and H. Kalt, “Co-existence of strongly and weakly localized random laser modes,” Nat. Photonics 3(5), 279–282 (2009). [CrossRef]
23. S. Mujumdar, V. Türck, R. Torre, and D. S. Wiersma, “Chaotic behavior of a random laser with static disorder,” Phys. Rev. A 76(3), 033807 (2007). [CrossRef]
27. A. Tulek, R. C. Polson, and Z. V. Vardeny, “Naturally occurring resonators in random lasing of π-conjugated polymer films,” Nat. Phys. 6(4), 303–310 (2010). [CrossRef]
28. R. C. Polson, M. E. Raikh, and Z. V. Vardeny, “Random lasing from weakly scattering media; spectrum universality in DOO–PPV polymer films,” Physica E 13(2-4), 1240–1242 (2002). [CrossRef]