LETTER pubs.acs.org/NanoLett
Random Laser Based on Waveguided Plasmonic Gain Channels Tianrui Zhai, Xinping Zhang,* Zhaoguang Pang, Xueqiong Su, Hongmei Liu, Shengfei Feng, and Li Wang Institute of Information Photonics Technology and College of Applied Sciences, Beijing University of Technology, Beijing 100124, China ABSTRACT: A waveguide-plasmonic scheme is constructed by coating the matrix of randomly distributed gold nanoisland structures with a layer of dye-doped polymer, which provides strong feedback or gain channels for the emission from the dye molecules and enables successful running of a random laser. Excellent overlap of the plasmonic resonance of the gold nanoislands with the photoluminescence spectrum of the dye molecules and the strong confinement mechanism provided by the active waveguide layer are the key essentials for the narrow-band and low-threshold operation of this random laser. This kind of feedback configuration potentially enables directional output from such random lasers. The flexible solution-processable fabrication of the plasmonic gold nanostructures not only enables easy realization of such a random laser but also provides mechanisms for the tuning and multicolor operation of the laser emission. KEYWORDS: Random laser, plasmonic scattering, waveguide, dye-doped polymer
R
andom lasers have been investigated extensively using a variety of materials and structures.1 5 In particular, the random lasing behaviors based on the localized surface plasmon resonance of metallic nanoparticles attracted broad interest.6 9 Dice et al. demonstrated surface-plasmon-enhanced random lasing from a suspension of silver nanoparticles in the solution of Rhodamine 6G in methanol.6 Popov et al. studied the enhancement effect of gold nanoparticles on the lasing characteristics in a polymer film doped with Rhodamine 6G.7 Meng et al. reported on observations of random lasers with the coherent feedback provided by the silver/metallic-dielectric core shell nanoparticles doped in a highly transparent polymer film.8,9 Plasmonic scattering by the metallic nanoparticles10 13 is the main mechanism of this kind of random lasers. A distributedfeedback polymer laser has been demonstrated using a twodimensional Bragg grating made of gold.14 A random laser scheme consisting of polymer films doped with liquid crystal droplets is proposed with directional and multiple-scattering feedback mechanisms.15 However, localized surface plasmon resonance of the gold nanostructures is not the mechanism that supports the lasing action. In this paper, a feedback scheme consisting of plasmonic scattering and waveguide confinement mechanisms16,17 provides effective gain channels and a random laser is achieved in the red spectral range, where the plasmonic resonance should overlap the emission of the active medium in spectrum as largely as possible. The waveguide is actually the layer of the gain medium, which is produced by spin coating the polymer of poly(methyl methacrylate) (PMMA) doped with a red dye of 4-(dicyanomethylene)-2-tert-butyl-6-(1,1,7,7-tetramethyljulolidin-4-yl-vinyl)-4H-pyran (DCJTB) onto the gold nanoisland structures. This kind of active waveguide not only provides highquality confinement of the radiation for efficient amplification16 r 2011 American Chemical Society
but also enables possible directional output of this kind of random laser. Panels a c of Figure 1 show the fabrication procedures of the random laser device. The colloidal solution of gold nanoparticles with a concentration of 100 mg/mL is first spin-coated onto the silica substrate (Figure 1a) before an annealing process at about 500 °C for 10 min. Thus, gold nanoisland structures (Figure 1b) with an area of 20 20 mm2 are produced. Then the blend solution of DCJTB and PMMA is spin-coated onto the silica substrate at a speed of 2000 rpm, which is prepared by mixing the solution of PMMA in chloroform with an concentration of 14 mg/mL and that of DCJTB in chloroform with a concentration of 3 mg/mL at a ratio (v/v) of 1:1. Thus, a PMMA waveguide doped with DCJTB is produced on top of the gold nanoisland structures, as shown in Figure 1c. Figure 1d illustrates the feedback scheme with the pumping (λP), the scattering, and the emitting (λE) picture, which forms the waveguided plasmonic feedback mechanism and supports stimulated emission and amplification of the dye radiation. The radiation from the dye molecules is scattered strongly multiple times by the gold nanoislands due to the particle plasmon resonance of the gold nanoislands, where the spectrum of particle plasmon resonance overlaps the emission spectrum of the dye, as will be shown in Figure 3. A large part of the scattered light may be reflected back totally at the PMMA/air interface to propagate within the active waveguide and scattered further by the gold nanostrucutres, which experiences strong amplification through stimulated emission. Received: July 7, 2011 Revised: August 30, 2011 Published: August 31, 2011 4295
dx.doi.org/10.1021/nl2023096 | Nano Lett. 2011, 11, 4295–4298
Nano Letters
LETTER
Figure 3. The extinction and the photoluminescence spectra of DCJTB and DCJTB on a gold nanoisland structure, respectively. The DCJTB (3 mg/mL) is doped in the PMMA (14 mg/mL); the solvent is chloroform. The blue line denotes the extinction spectrum of the gold nanoisland structures.
Figure 1. The schematic illustration of the fabrication process and basic principles of the random laser device based on the waveguided plasmonic gold nanostructures. Figure 4. The PL lifetime of DCJTB on the silica substrate and the gold nanoisland structure.
Figure 2. AFM height image of the gold nanoisland structures prepared on a silica substrate. The lower panel shows the cross section of the gold nanoislands at the position marked out by the green line.
Figure 2 shows the atomic force microscopic (AFM) images of the gold nanoisland structures fabricated on a silica substrate. As has been described in our previous work,18,19 at an annealing temperature of above 350 °C, no continuous gold film can be
produced; instead, gold nanoisland structures with considerably homogeneous shape and size with random distribution can be produced due to the large surface tension of the molten gold. Both the AFM height image and the profile characterization in Figure 2 show that the gold nanoislands have a mean height of about 80 nm and a mean diameter of about 150 nm. The separation between the gold nanoislands is in almost the same scale as the size of the gold nanoislands. The random laser device is pumped by a frequency doubled neodymium-doped yttrium aluminum garnet pulsed laser at a wavelength of 532 nm, a repetition rate of 10 Hz, a pulse length of 30 ns, and a pulse energy of up to 50 mJ. Figure 3 shows that DCJTP doped in PMMA has an absorption spectrum (black curve) centered at about 505 nm and a photoluminescence (PL) spectrum (red curve) centered at about 630 nm; the extinction spectrum of the gold nanostructures is centered at 600 nm in air (blue curve) and at 640 nm (green curve) when coated with the active waveguide layer of the DCJTB doped PMMA. The PL spectrum of DCJTB becomes broadened and red-shifted to about 650 nm when the active layer of DCJTB doped PMMA is coated onto the gold nanoisland structures, which results clearly from the strong scattering by the gold nanoislands as the PL spectrum overlaps that of particle plasmon resonance of the gold nanostructures. Thus, particle plasmon resonance of the gold islands has induced selective enhancement of the PL spectrum, and this is the basic mechanism for the stimulated 4296
dx.doi.org/10.1021/nl2023096 |Nano Lett. 2011, 11, 4295–4298
Nano Letters
LETTER
Figure 5. Measured spectra of the random lasing emission at different pump fluences. The right upper inset denotes the output intensity of the random laser as a function of the pump fluence, indicating a pump threshold of about 0.1 MW/cm2. The right lower inset denotes the laser mode around the pump threshold.
emission and the random laser radiation. This can be confirmed by the comparison between the PL decay dynamics of DCJTB with (solid yellow circles) and without (solid red circles) gold nanoislands, as shown in Figure 4. Simple analysis found that the interaction with particle plasmon resonance of the gold nanostructures reduced the PL lifetime from about 1.8 to 1.3 ns. Therefore, the interaction with the gold nanostructures has led to the reduction in the PL lifetime and the enhanced PL intensity of the dye molecules, implying possible amplification of the fluorescence through stimulated emission. The random lasing results from multiple events of light scattering taking place sequentially by the disordered plasmonic gold nanostructures at their interfaces with the PMMA doped with DCJTB, which is enhanced or multiplied by the PMMA layer through the confinement of the scattered light into the waveguide. This can be equalized to a kind of microcavity effect, where the “round-trip” forms between the total reflection at the PMMA air interface and the strong scattering by the gold nanoislands. Thus, each “round-trip” corresponds to a gain process through a “scattering total reflection scattering” process and the lasing action depends strongly on the minimum mean free path length, which is defined as lmin = 2n2 d/(n2 1)1/2, where n and d are the refractive index and the thickness of the gain medium, respectively. This path length is multiplied by the total reflection process and extended within the waveguide. In our experiments, n ≈ 1.49, d ≈ 300 nm, and we obtain lmin = 1.2 μm. Obviously, this minimum mean free path length requires that the separation between the gold nanoislands be smaller than S = lmin sin θC, where θC = sin 1(1/n) is the critical angle for total reflection at the PMMA/air interface. Thus, we obtain S ≈ 0.8 μm. If looking at the AFM image in Figure 2, as well as those in Figure 6, the separation between the gold nanoislands is always smaller than the value of S. Therefore, the unpatterned gold nanostructures fabricated using solution-processable gold nanoparticles facilitate satisfying spatial distribution and geometric arrangement for random lasers. Figure 5 shows the spectra of the random laser emission at different pump fluences measured using a spectrometer of Maya 2000 PRO from Ocean Optics. The laser emission is centered at about 656 nm and has a typical line width of about 10 nm at full width at half-maximum (fwhm). Thus, the minimum mean free path length (lmin) approximately equals 1.83λ0, where λ0 is the center wavelength of the random laser emission. According to
Figure 6. Tuning the emission wavelength of the random laser by changing the size of the gold nanoisland with the AFM images of the corresponding gold nanoisland structures shown in the top panel.
the definitions of the free path length of random lasers and the experimental work reported before,20 23 this is an excellent relationship to support the lasing actions in such a random device. The inset on the right-top corner of Figure 5 shows the output laser intensity as a function of the pump fluence, implying a pump threshold of about 0.1 MW/cm2. The inset on the rightbottom corner of Figure 5 shows the enlarged spectrum of the single-mode oscillation of the laser when the pump fluence is around the threshold, which has a bandwidth narrower than 0.5 nm at fwhm, where the spectrometer has a resolution of 0.2 nm. This implies that the broad-band emission of the random laser actually contains a number of oscillation modes. The spectral position of the random lasing depends strongly on that of particle plasmon resonance of the gold nanoislands, which can be confirmed by the measurements in Figure 6. The top panel of Figure 6 shows the AFM images of the gold nanoisland structures before they are spin-coated with DCJTPdoped PMMA and transformed into a random laser device. According to the statistical measurements, the gold nanoislands in device A have a mean diameter of about 90 nm and those in B have a mean diameter of about 160 nm. This kind of increase in the size of the gold nanoislands actually tuned the particle plasmon resonance to the red. As shown by the dashed curves in the bottom panel of Figure 6, the spectral peak of particle plasmon resonance shifts from about 625 to 655 nm if sample B is used to replace sample A in the fabrication of the random laser device. As a result, the lasing spectrum shifts from about 642 to 659 nm. This not only shows the tunability of the random laser through changing the size of the gold nanoislands and consequently tuning the spectrum of particle plasmon resonance but also confirms that the dominant mechanism for the random laser is the scattering by the gold nanoislands through particle plasmon resonance. The random laser can be easily tuned by changing the size of the gold nanoislands, which can be achieved by modifying the concentration of the colloidal solution of the gold nanoparticle or the annealing temperature during the fabrication process.19 Furthermore, hybrid structures with different sizes of gold nanostructures can be fabricated on a same substrate using a solution-processable method, so that the spectrum of particle 4297
dx.doi.org/10.1021/nl2023096 |Nano Lett. 2011, 11, 4295–4298
Nano Letters plasmon resonance can be largely broadened or be designed to take place at multiple spectral positions. Thus, if the active medium is a mixture of different organic semiconductors that emit different wavelengths, a multicolor or even white light random laser may be possibly achieved. However, the energy transfer process between the organic molecules should be taken into account in designing such lasers, which may quench the emission of one or more component materials. The waveguide-plasmon-feedback scheme not only reduces the pump threshold due to further confinement of the fluorescence by the waveguide to the gain medium as compared with the bodily embedded configuration6 9 but also implies possibly directional output of this a random laser. This is a special feature and an advantage of this kind of random laser design. Additionally, it should be noted that the randomly distributed gold nanoisland structures actually apply a spatial modulation of the active layer of DCJTB-doped PMMA, the intrinsic scattering at the organic metal interfaces has also enhanced the scattering mechanisms and provides an additional feed-back mechanism for the random lasing process. In conclusion, we demonstrated a new kind of random laser scheme, which consists of a bottom layer of randomly distributed gold nanoislands and a top layer of active waveguide made of DCJTB doped PMMA. The strong plasmonic scattering by the gold nanostructures and the strong confinement by the active waveguide extend free path length significantly, enabling low pump threshold and high conversion efficiency of the random laser. In particular, the “waveguide-plasmonic scattering” mechanism potentially facilitates directional or utilizable output of the random laser. Furthermore, this kind of random laser can be tuned by changing the size of gold nanoisland structures through controlling the annealing temperature or the concentration of the colloidal concentration during the solution-processed fabrication technique.
LETTER
(7) Popov, O.; Zilbershtein, A.; Davidov, D. Appl. Phys. Lett. 2006, 89, 191116. (8) Meng, X.; Fujita, K.; Zong, Y.; Murai, S.; Tanaka, K. Appl. Phys. Lett. 2008, 92, 201112. (9) Meng, X.; Fujita, K.; Murai, S.; Matoba, T.; Tanaka, K. Nano Lett. 2011, 11, 1374–1378. (10) Gopinath, A.; Boriskina, S. V.; Feng, N. N.; Reinhard, B. M.; Negro, L. D. Nano Lett. 2008, 8, 2423–2431. (11) Tam, F.; Goodrich, G. P.; Johnson, B. R.; Halas, N. J. Nano Lett. 2007, 7, 496–501. (12) Zhou, Z. K.; Peng, X. N.; Yang, Z. J.; Zhang, Z. S.; Li, M.; Su, X. R.; Zhang, Q.; Shan, X.; Wang, Q. Q.; Zhang, Z. Nano Lett. 2011, 11, 49–55. (13) Catchpole, K.; Polman, A. Opt. Express 2008, 16, 21793–21800. (14) Stehr, J.; Crewett, J.; Schindler, F.; Sperling, R.; von Plessen, G.; Lemmer, U.; Lupton, J. M.; Klar, T. A.; Feldmann, J.; Holleitner, A. W.; Forster, M.; Scherf, U. Adv. Mater. 2003, 15, 1726. (15) Gottardo, S.; Cavalieri, S.; Yaroshchuk, O.; Wiersma, D. Phys. Rev. Lett. 2004, 93, 263901. (16) Zhai, T.; Zhang, X.; Pang, Z. Opt. Express 2011, 19, 6487–6492. (17) Zhai, T.; Zhang, X.; Pang, Z.; Dou, F. Adv. Mater. 2011, 23, 1860–1864. (18) Liu, H.; Zhang, X. Photonics Nanostruct.: Fundam. Appl. 2010, 8, 131. (19) Zhang, X.; Liu, H.; Feng, S. Nanotechnology 2009, 20, 425303. (20) Cao, H.; Xu, J.; Chang, S. H.; Ho, S. Phys. Rev. E 2000, 61, 1985. (21) Wiersma, D. Nat. Phys. 2008, 4, 359–367. (22) Wu, X.; Fang, W.; Yamilov, A.; Chabanov, A.; Asatryan, A.; Botten, L.; Cao, H. Phys. Rev. A 2006, 74, 53812. (23) Mujumdar, S.; Ricci, M.; Torre, R.; Wiersma, D. Phys. Rev. Lett. 2004, 93, 53903.
’ AUTHOR INFORMATION Corresponding Author
*E-mail:
[email protected].
’ ACKNOWLEDGMENT The authors acknowledge the National Natural Science Foundation of China (11074018), the Program for New Century Excellent Talents in University (NCET), the Research Fund for the Doctoral Program of Higher Education of China (RFDP, 20091103110012), and the Doctoral Scientific Research Foundation of Beijing University of Technology (X0006111201103) for the financial support. ’ REFERENCES (1) Lawandy, N.; Balachandran, R.; Gomes, A.; Sauvain, E. Nature 1994, 368, 436–438. (2) Frolov, S.; Gellermann, W.; Ozaki, M.; Yoshino, K.; Vardeny, Z. Phys. Rev. Lett. 1997, 78, 729–732. (3) Cao, H.; Zhao, Y.; Ho, S.; Seelig, E.; Wang, Q.; Chang, R. Phys. Rev. Lett. 1999, 82, 2278–2281. (4) Wiersma, D. Nat. Phys. 2008, 4, 359–367. (5) Zhai, T.; Zhou, Y.; Chen, S.; Wang, Z.; Shi, J.; Liu, D.; Zhang, X. Phys. Rev. A 2010, 82, 023824. (6) Dice, G.; Mujumdar, S.; Elezzabi, A. Appl. Phys. Lett. 2005, 86, 131105. 4298
dx.doi.org/10.1021/nl2023096 |Nano Lett. 2011, 11, 4295–4298