Open Access
Add to BibSonomyAbstract
Optically pumped organic thin-film lasers were fabricated by stacking top and bottom Bragg reflectors with an inner-cavity active layer between the reflectors. We observed single-longitudinal-mode laser operation at 555 nm by doping pyrromethene-567 dye into all the stacked layers. The threshold of the laser was 30% less than that of a laser operated with nondoped Bragg reflectors, which was in good agreement with calculations.
©2006 Optical Society of America
1. Introduction
The development of compact lasers covering the visible wavelength spectrum has attracted interest in the fields of spectroscopy, sensing, and environmental and biomedical applications. Solid-state organic lasers have great potential for these applications owing to many varieties of organic dye, their spectrally broad gain region, and their low production costs. To make the best use of such features, the miniaturization of the organic laser system is indispensable. The strategy that should be taken first is to decrease the lasing threshold. Ideal photon confinement can be attained using three-dimensional (3D) photonic band gap structures; however, it is very difficult to fabricate a 3D structure operating in the visible spectral range. Thus, vertical-cavity surface-emitting lasers with a distributed Bragg reflector (DBR) have been fabricated using evaporated inorganic thin films or spin-coated organic thin films to provide a complete 1D feedback [1–4]. However, the absorption of the pump beam and the optical gain-length product are limited owing to the short active regions of the vertical-cavity structures.
In this study, we demonstrate a polymer-based vertical-cavity DBR laser by doping dye into the DBR mirrors as well as the inner-cavity active layer. We examine the spectral characteristics and threshold of the laser in comparison with a laser fabricated with nondoped DBR mirrors.
2. Device principles
In our vertical-cavity organic laser scheme, the top and bottom DBR mirrors consist of quarter-wave layers alternately coated of high- and low-refractive-index polymers. The inner-cavity active layer is inserted between the two DBR mirrors to satisfy the round-trip phase matching condition. We examine two types of DBR mirror, namely, the active DBR mirror and the passive DBR mirror, which differ in that the quarter-wave layers either contain or do not contain dye molecules.
Fig. 1. Schematic diagrams of vertical-cavity organic lasers: (a) passive DBR laser and (b) active DBR laser.
Figure 1 shows schematics of the vertical-cavity organic lasers that have either the passive DBR mirrors or the active DBR mirrors. All the layers of the active DBR laser are doped with laser dye to enable the efficient absorption of pump light and supply the optical gain for the whole layer. We calculated the transmission spectra of the laser cavity by the transfer matrix method (TMM) and analyzed the threshold gain [5]. We assumed refractive indices of 1.47 for a low-refractive-index layer and 1.68 for a high-refractive-index layer; these refractive indices are derived from cellulose acetate (CA) and polyvinylcarbazole (PVK), respectively [6]. Figure 2 shows the calculated results of the variations in threshold gain with respect to the pair number of the quarter-wave layers in the bottom DBR mirror. We assume inner-cavity active layers with different thicknesses of 1λ and 9λ, where λ denotes the laser wavelength. Note that the active DBR mirrors give lower threshold gains than the passive DBR mirrors. The lower threshold gains are achieved by the distributed feedback gain effect in the whole dye-doped layer. In the case of the 1λ cavity, the effect of adding gain to the DBR mirrors is very substantial, so that the passive DBR laser requires five more pairs in each of the top and bottom DBR mirrors than the active DBR laser to attain an equivalent threshold gain. For the 9λ cavity, the difference in threshold gain between the two types of laser is reduced owing to an increase in the proportion of the inner-cavity volume in the whole layer. Although the effect of distributed gain will be much more pronounced for the 1λ cavity as compared to the 9λ cavity, we deliberately choose to study the effect in the 9λ cavity because that cavity is more easily brought above threshold for the two types of laser, and the 1λ cavity is, for experimental reasons, not feasible. Thus we can observe the effect of the active DBR mirrors by the comparison with the passive DBR mirrors.
Fig. 2. Threshold gains as functions of pair number of high- and low-refractive-index layers in bottom DBR mirror with inner-cavity layer thicknesses of (a) 1λ and (b) 9λ. The top DBR mirror has 1.5 fewer pairs than the bottom DBR mirror.
3. Experimental method
The laser structures were realized by alternately spin-coating a diacetone alcohol solution of CA and a chlorobenzene solution of PVK on silicon substrate [3]. We found that the pyrromethene-567 (P567) dye was dissolvable both in diacetone alcohol and in chlorobenzene. We utilized P567 as a doping dye with a concentration of 1% by weight. The top and bottom DBR mirrors were composed of 12.5 pairs and 14 pairs of CA/PVK quarter-wave layers, respectively. An inner-cavity CA layer of ~3.6 µm thickness was inserted between the two DBR mirrors. In order to avoid the dye deterioration, a baking process of the polymer films should be done at as low temperature as possible. At a baking temperature of 120 °C or less, the surface of the film became rough when the CA/PVK/CA layers were sequentially coated onto the substrate. This is attributable to that the diacetone alcohol penetrates to the lower CA layer through the intermediate PVK layer. The PVK solvent was dried enough by adjusting a baking temperature to 130 °C, and a smooth surface was obtained.
In an optical pumping experiment, the vertical-cavity laser structures were irradiated by a frequency-doubled Nd:YAG laser system operating at 532 nm with a 7 ns pulse width and a 10 Hz repetition rate. The pump beam was focused onto the structure to obtain a spot ~600 µm in diameter. The emitted light was detected perpendicular to the sample surface using a monochromator in combination with a charge-coupled-device array detector, providing a spectral resolution of 0.22 nm.
4. Results and discussion
Figure 3 shows the photoluminescence (PL) spectra of the passive and the active DBR lasers at low excitation intensity, showing a broad spectral range from 540 nm to 610 nm. The PL spectra were modified by the cavity modes and the mode peaks arose from the spontaneous emission background. By applying the calculated transmission spectrum of the passive vertical-cavity structure to the experimental results, we determined the positions of all peaks, namely, of the two peaks at the stopband edges and of the three peaks inside the stopband. The stopband of the resonator is characterized by the center wavelength of about 585–590 nm and a full width at half maximum (FWHM) of 80 nm. We clearly observed longitudinal mode peaks at almost equal intervals inside the stopband in the case of the passive DBR laser. The cavity mode peaks match reasonably well with those of the calculated transmission spectrum. Compared with this result, the cavity mode spacing of the active DBR laser structure was slightly shortened. This can be attributed to the difference in the effective length between the passive and the active DBR mirrors. Since the effective DBR length is in inverse proportion to the coupling coefficient [7], the refractive index difference between the two doped materials would be decreased compared to the nondoped conditions.
Fig. 3. PL spectra below lasing threshold: (a) passive DBR laser and (b) active DBR laser. The dotted curves represent the calculated transmission spectra of the passive vertical-cavity structure.
Figure 4 shows the dependence of the emission intensity of the passive DBR laser on pump pulse energy. We obtained a lasing action at a threshold of ~88 µJ/pulse, which corresponds to a pulse energy density of ~31 mJ/cm2. Above the threshold, the additional energy supplied by the external pumping laser reduces the mode competition to a single longitudinal mode, driving the emission into a highly resolved peak at 585 nm, with a FWHM linewidth narrower than 0.7 nm. Figure 5 shows emission intensity as a function of pump pulse energy for the active DBR laser. We obtained lasing from the active DBR laser at a threshold of ~60 µJ/pulse, which corresponds to a pulse energy density of ~21 mJ/cm2. We observed a decrease in the lasing threshold of about 30% compared with that of the same laser structure with nondoped DBR mirrors. Since the inner-cavity thickness of ~3.6 µm roughly corresponds to 9λ, we can compare the measured results with the calculated results in Fig. 2(b). The measured result of the 30% reduction is in a good agreement with the calculated result for 14 pair layers of the bottom DBR mirror. The active DBR laser has the advantage of absorbing pump light in smaller volumes, owing to the comparatively large absorption length and the resulting larger gain-length product. The inset of Fig. 5 shows the lasing spectrum with a center wavelength of 555 nm. The linewidth was measured to be about 1.2 nm, which was larger than that for the passive DBR laser. This broadening of the linewidth is considered to be the influence of the amplified spontaneous emission in the dye-doped DBR mirrors.
Fig. 4. Input-output characteristics of passive DBR laser. The inset shows the lasing spectrum at 110 µJ/pulse.
Fig. 5. Input-output characteristics of active DBR laser. The inset shows the lasing spectrum at 80 µJ/pulse.
5. Conclusions
We have demonstrated an optically pumped organic vertical-cavity laser, in which the whole layer including the Bragg reflectors was doped with laser dye. The lasing threshold was reduced compared with a laser with nondoped Bragg reflectors under similar conditions. A simple TMM model provided insight into the spectral and threshold behaviors of the lasers. These results imply a major reduction in threshold for realistic 1λ cavities. Further improvements to the threshold are being pursued by adjusting the dye concentration and rearranging the stacked layers.
References and links
1. V. Bulović, V. G. Kozlov, V. B. Khalfin, and S. R. Forrest, “Transform-limited, narrow-linewidth lasing action in organic semiconductor microcavities,” Science 279, 553–555 (1998). [CrossRef] [PubMed]
2. A. Arena, S. Patanè, G. Saitta, G. Rizzo, S. Galvagno, and G. Neri, “Photoluminescence from organic-inorganic multilayers based on sol-gel derived titania,” J. Non-Cryst. Solids 331, 263–268 (2003). [CrossRef]
3. T. Komikado, A. Inoue, K. Masuda, T. Ando, and S. Umegaki, “A surface-emitting distributed-feedback dye laser fabricated by spin-coating organic polymers,” in Proceedings of Conference on Lasers and Electro-Optics, 3 (Optical Society of America, Baltimore, 2005), pp. 2016–2018.
4. L. Persano, A. Camposeo, P. D. Carro, E. Mele, R. Cingolani, and D. Pisignano, “Very high-quality distributed Bragg reflectors for organic lasing applications by reactive electron-beam deposition,” Opt. Express 14, 1951–1956 (2006). [CrossRef] [PubMed]
5. M. Yamada and K. Sakuda, “Analysis of almost-periodic distributed feedback slab waveguides via a fundamental matrix approach,” Appl. Opt. 26, 3474–3478 (1987). [CrossRef] [PubMed]
6. J. Brandrup, E. H. Immergut, and E. A. Grulke, eds., Polymer Handbook, 4th ed. (John Wiley, New York, 1999), VI–571.
7. A. Yariv, Optical Electronics in Modern Communications, 5th ed. (Oxford University Press, New York, 1997), Chap. 16.
